The Line Below the Invoice
Haute Lumière · The Price Nobody Charged · Chapter One
Chapter One
The price a firm receives is settled below the invoice, in a dozen places nobody adds up.
The Line Below the Invoice
Haute Lumière · The Price Nobody Charged · Chapter One
It is ten to six on a Tuesday in February, and the commercial analyst at a distributor of plumbing and heating supplies has a year of one customer's transactions spread across two desks. She was asked on Monday a question that sounded simple: why has the margin on the firm's third-largest account fallen, when the price list has gone up twice? The account is the one the sales director mentions at every conference. Its buyer sends a card at Christmas.
So she has printed everything. The invoices first, each one at the agreed discount from list, which is what everyone in the building means when they talk about this customer's price. Then the other paper, the kind that never reaches an invoice. The credit notes for stock returned unopened, refunded in full. The year-end volume rebate, paid in January, although the customer's volume fell below the threshold the rebate was written for. The early-payment discount, taken on every invoice, by a customer whose payments arrive on average at fifty-eight days. The freight on every order, including the half-pallets sent by van the same afternoon. A marketing contribution for a trade show the customer did not attend.
She adds it up for the product the customer buys most, a thermostatic valve. Then she does the same for a two-person installer in the next town, who buys four a month and pays on the day. The installer pays more for the valve. Not a little more. Her stomach tightens in the way it does before a meeting she would rather not have.
She is a composite, drawn from the way this afternoon is described in the pricing literature below and in the house's own practice, and no firm is meant. What she has found is ordinary, and it is ordinary in nearly every firm that sells to other firms.
Ammanuel Santa Anna's High Frequency Workbook is written for a person's own money, not a company's.1 One of its sentences reads, on the analyst's two desks, as if it had been written for a finance director:
Financial avoidance is the money equivalent of emotional suppression. It feels protective in the short term and creates compounding damage in the long term. Every day you do not know your numbers is a day you are making financial decisions in the dark.
This chapter makes one claim. The price a firm actually receives is decided in many places after the price is set, by people who each see one line and never the total. The leak below the invoice is rarely anyone's decision. It is the sum of reasonable concessions nobody revisits, and because nobody owns the sum, nobody sees it. Making it visible is the first and cheapest way a firm finds revenue it has already earned.
In 1992 two McKinsey consultants, Michael Marn and Robert Rosiello, published an article in the Harvard Business Review under the title Managing Price, Gaining Profit.2 Its opening sentence is still the shortest argument for the subject: "The fastest and most effective way for a company to realize its maximum profit is to get its pricing right." The article is best remembered for two pictures, and both have been taught in pricing courses ever since.
The first is the pocket price waterfall. It begins with the list price and steps down, bar by bar, to the invoice price, through the discounts printed on the invoice. Then it keeps going. Below the invoice it subtracts everything else the transaction costs the seller: payment-term discounts, annual rebates, promotional allowances, freight paid on the customer's behalf, the cost of returns. What is left is the pocket price, the money that actually stays in the seller's pocket. The point of the picture is that most firms manage the top half of it, because the top half is on the invoice, and very few manage the bottom half, because the bottom half is scattered across ledgers that different departments keep.
The second picture is the pocket price band. Take one product, line up every customer who bought it, and sort them by pocket price. The band shows how wide the spread is. In the cases the article and its successors describe, it is often far wider than anyone in the firm expected, and it bears little relation to the size of the customer or the cost of serving it.
The article's own text was not reopened for this chapter beyond its opening; the account of the two pictures here is the one that has been taught from it, and the figures it gave for particular companies are not repeated. What can be done is the arithmetic that explains why the subject matters, and it is worth doing once, by hand. Suppose a firm earns an operating margin of eight per cent. On every hundred pounds of sales it keeps eight. If the price it actually receives rises by one per cent, and nothing else changes, it keeps nine. One pound on eight is a rise of a little over twelve per cent in operating profit. The same firm would need to sell many more units to make the same pound, because each new unit also brings its costs. That is the whole of the arithmetic, and it runs in both directions. A leak of one per cent below the invoice is an eighth of the profit, gone without a meeting.
Santa Anna's canon on value-stream mapping has a phrase for the kind of flow the bottom half of the waterfall describes. It writes of shadow streams, "value flows that exist below the surface of official process, unacknowledged by leadership, unmapped by conventional assessment, but essential to the system's actual functioning." The canon means the quiet work that keeps a firm alive, the mediating and mentoring that appear on no chart. The waterfall is a shadow stream running the other way. It is just as unmapped, just as essential to the system's actual functioning, and it carries money out rather than in.
If the leak is so large and so simple to see, why does it persist? The most careful answer comes from three researchers who spent time inside one company. In 2003 Shantanu Dutta, Mark Zbaracki and Mark Bergen published a study of the pricing process of a large manufacturing firm in the American Midwest.3 Their argument begins with a distinction that most strategy writing skips. A firm can create value, they wrote, and still fail to keep it: "Even a firm that has created value, however, may not capture the potential rents associated with that value." To capture it, a firm must set and hold the right prices, and most theories simply assume it can.
They argued the opposite: "we argue that pricing is a capability." Setting the right price, in their account, takes resources, routines and skills, built over time and spread across people, and a firm that lacks them cannot price well however good its products are. They show the capability at work in one company, in the detail of who gathers what information, who decides, and how a decision travels to the customer. It is one firm, studied closely, and its strength is depth rather than reach. Its central point does not depend on the firm. Price is not a number someone types into a system. It is the output of an organisation.
The study also makes a point that is easy to miss, and it matters for anyone who thinks of recovering price as a contest with the customer. Firms without good pricing processes, the authors wrote, "may be unable to set prices that reflect the wishes of its customers, so the customers may misuse their resources." A price that is wrong in either direction misleads the customer as well as the seller. A customer who is charged too little for same-day delivery uses it for orders that could have waited, and the firm's vans run half empty to serve a habit the price created. A customer charged too much for a product that costs little to serve buys less of it than it would have. Pricing well is not only how a firm keeps what it creates. It is how it tells its customers the truth about what things cost.
Read the analyst's two desks through that sentence. The invoice discount was set by the account manager, who is measured on volume. The rebate was written by the sales director three years ago, to win the account from a competitor, and is paid by finance, who pay what the contract says without asking whether the volume was met. The early-payment discount is applied by the system automatically; credit control sees the late payments, but its job is to collect the money, not to reclaim a discount. Free delivery was agreed when the customer's yard was two miles from the depot, before the customer moved. The returns policy belongs to the warehouse, which is measured on accuracy, and an unopened box is the easiest return there is.
Each of these people is competent. Each of them owns one bar of the waterfall. None of them owns the pocket price. The leak lives in the space between their jobs, and nobody is measured on a space.
The house's Luminous 100, its reading of the frameworks every consultancy sells, says of Porter's value chain that "The competitive advantage lives in specific activities, not the whole company." The pocket price is the same truth seen from the bottom of the account. The loss does not live in "the customer" or "the market". It lives in specific lines, kept by specific desks, and it can only be found there.
Every bar had a reason
Haute Lumière · The Price Nobody Charged · Chapter One
It is worth being fair to the waterfall, because a leader who sees it for the first time is tempted to treat every bar as a mistake. Almost none of them began as one.
The rebate won an account the firm wanted, against a competitor who was offering more. The early-payment discount was agreed when the customer paid in ten days, and it bought the firm cash it needed that year. The free delivery made sense when the van passed the customer's door on its way to somewhere else. The full credit for unopened returns kept a customer who had over-ordered for a project that was cancelled. A marketing contribution is how distributors in this trade show their suppliers they are serious. If the analyst had been in each room when each concession was made, she would probably have agreed to it.
The trouble is not the concession. It is that the reason expires and the concession does not. The customer starts paying at fifty-eight days and keeps the discount meant for ten. The volume falls and the rebate is paid anyway, because the clause that ties it to volume is in a contract nobody in finance has read since it was signed. The customer moves its yard and the van makes a special trip. Each bar outlives the reason it was granted for, and nobody is asked to notice, because each bar is somebody else's.
Santa Anna's Ecstatic Economics puts the principle underneath in a sentence that was written about currencies and holds just as well for a rebate: "money is a design choice, not a law of nature." Every bar below the invoice is a design choice. The firm made it once, on purpose, for a reason. What it has not done is make it again, on purpose, now that the reason has changed. A concession that is chosen every year is a decision. A concession that is simply renewed is a leak with a history.
It would be easy to tell this story as if the account manager were the cause, and it would be wrong. His view of the same year is worth hearing, because it is the view from which most of the waterfall is built.
He has looked after the account for four years. He knows the buyer's children's names and the date of the customer's busy season. He knows, because the buyer has told him more than once, that a competitor has offered a better price, though he has never seen the competitor's quotation in writing. Every concession he has agreed was agreed in a conversation in which the alternative seemed to be losing the account, and the account is a large part of the volume on which he is paid. When the customer asked to keep the early-payment discount while paying later, it was in the week a rival's representative had been seen in the customer's yard. When the customer asked for the same-day vans, it was after a delivery had gone wrong, and the vans felt like an apology.
None of these conversations was recorded anywhere a finance director would read. The firm's systems record what was invoiced. They do not record what was asked, by whom, under what pressure, or what the alternative was believed to be. So when the margin falls, finance sees a number and the account manager sees four years of reasonable decisions, and the two views never meet. Each side, privately, thinks the other does not understand the business.
The account manager also carries something the analyst does not, and it is in his body rather than in the ledgers. He is the one who will have to make the phone call. He is the one the buyer will be disappointed in. He has spent four years being the person at the firm who says yes to this customer, and he has, without ever deciding to, come to feel that saying yes is his job. Any change to the waterfall will be experienced by him, first, as a change to who he is to the buyer.
That is why the waterfall cannot be fixed by finance alone, and why it should never be presented to the account manager as a list of his mistakes. Every bar on it was a reasonable answer to a real question. What the firm lacked was not judgement in the moment. It was a place where the moment was written down, and a time when it was read again, which is the same lack that leaves a professional firm's extra work unbilled (Chapter 2).
There is a human reason the bars are rarely revisited, and it is the one the analyst feels in her stomach. To ask about the rebate is to ask about the sales director's best deal. To ask about the early-payment discount is to ask the account manager to ring a friend and say something unwelcome. The High Frequency Workbook, writing about why people avoid their own accounts, says that money "is a mirror. Not of your worth — never of your worth — but of your patterns." The waterfall is that mirror for an organisation. It does not show whether the firm is good. It shows what the firm has been unwilling to ask.
The analyst does not stop at one customer. On Wednesday morning she takes the thermostatic valve, the firm's most common line, and works out the pocket price for every customer who bought more than a handful in the year. There are a little over two hundred. She sorts them, and draws the band.
It is wide, and it is not shaped the way anyone would have designed it. Some of the largest customers are near the bottom, as expected, because they buy in volume and cost little to serve per unit. But some large customers are near the top, and some small ones are near the bottom, and the order has more to do with which account manager each customer has had, and how long ago its terms were last negotiated, than with anything the firm would defend in public. The installer in the next town is in the top tenth. He pays on the day, takes no rebate, collects his own stock, and returns nothing. He is, by any measure of cost to serve, one of the cheapest customers the firm has. He pays among the highest prices.
The analyst sits with the band for a while before she shows it to anyone. It is an uncomfortable picture, and it is tempting to read it as a verdict on the firm or on the people who built it. The High Frequency Workbook has a short sentence for the moment of first looking at numbers that have been avoided, written for a person facing her own accounts: "Whatever it is, it is just a number." The band is just a number, two hundred times over. It describes what has happened. It does not yet say what anyone should do, and the first thing it asks for is not blame but a second look.
That is the part of the band leaders tend to miss, and it changes what the exercise is for. The band is not only a map of revenue the firm has failed to collect. It is also a map of fairness. Somewhere on it are customers paying more than their business costs the firm, subsidising customers who pay less, and neither group has been told. A firm that reads the band only as a list of accounts to squeeze has read half of it. The other half is a list of customers it may be overcharging, and those customers are often the small, loyal, uncomplaining ones whose business would be the easiest to lose to a competitor who noticed first.
This matters for anyone who wants to share in what the band reveals. A gain that comes from charging a captive customer more than it agreed is not a gain anyone should want to measure. A gain that comes from honouring what was agreed, from tying each concession back to the reason it was given and from pricing the cheapest-to-serve customers as fairly as the most demanding ones, is a gain that the customer can see and survive. The difference between the two is the difference between a firm that recovers its revenue and one that merely moves cost onto the people least able to argue.
Drawing it by hand
Haute Lumière · The Price Nobody Charged · Chapter One
What the analyst does next is the practice this chapter is about, and it is more a way of looking than a technique.
On Thursday afternoon she asks the account manager to sit with her for an hour. She has a sheet of A3 paper and a pencil, and she draws the waterfall for the one customer by hand, bar by bar, from the list price down. Each bar comes from its own ledger, and she writes the source beside it: the invoice system for the invoice discount, the rebate file for the rebate, the carrier's statements for the freight, the warehouse log for the returns. Drawing it by hand is slower than a spreadsheet, and that is the point. The account manager watches the bars appear one at a time, each one familiar, and sees for the first time how far down the page the last one ends.
Then they go back up the page, bar by bar, and ask three questions of each. Who approved this, and when? What was the reason at the time? Does the reason still hold? They write the answers on the sheet. The invoice discount holds; the customer's volume still earns it. The rebate's reason does not hold, because the volume it was written for has not been reached for two years. The early-payment discount's reason does not hold either. The free delivery is mixed: it holds for full pallets on the regular route and not for the same-day vans. The returns credit holds for a project cancellation and not for routine over-ordering.
The account manager is quiet for most of this. Then he says something the analyst did not expect. He has known about the late payments for a year. He did not raise them, because credit control had not raised them, and he assumed that if it mattered someone in finance would say so. Credit control, it turns out, assumed the discount was the account manager's business. Readers of Where the Flow Stops will know the pattern. It is the meeting after the meeting, held between two departments that never met.
The decisions they take are small, and they are taken in order. First, every bar gets one of three answers, written beside it: keep, because the reason holds; renegotiate, openly, with the customer; or stop. Second, nothing is stopped by surprise. The account manager rings the customer's buyer the following week, before any invoice changes, and says plainly what the firm has found and what it proposes: the early-payment discount for payments within its terms, the rebate tied to the volume it was written for, free delivery on full pallets and a charge for same-day vans. The house has a book line on stating money plainly, from its own guide to selling, and the advice transfers exactly: "Present the investment as a clear, confident statement — not an apology". Third, the analyst puts the installer's price on the list too, and the firm brings it down to where his business belongs.
The phone call goes better than the account manager feared, and worse than he hoped. The buyer agrees to the delivery change at once, argues for a month about the rebate, and accepts the payment terms only after a meeting. What makes it possible to have the conversation at all is that the firm can show, on one sheet, exactly what it is asking and why, and that some of what it has found is in the customer's favour. The harder part, renegotiating a term that has been ignored for years without treating the customer as the one at fault, is the conversation about the clause in the drawer (Chapter 3), and it has its own discipline.
There are predictable ways for this to go wrong. A firm that draws the waterfall and then announces new terms by letter has turned a discovery into an ambush. A firm that draws it for its largest customers only, and not for the band, will recover from the accounts strong enough to push back and miss the ones it is overcharging. And a firm that leaves the incentives unchanged will find the bars growing back, because the account manager is still measured on volume and the cheapest way to volume is a concession nobody sees. Why that is so, and what the last week of a quarter does to the price (Chapter 5), is a matter the research on sales incentives has been precise about. Chapter 7 carries the run sheet for the waterfall and the band, the sources for each bar, the three questions and the order of the conversations, and Chapter 6 places it in the order of a first ninety days.
The gain in all of this is real and it is worth naming exactly without inventing its size. It is the difference between the pocket price the firm receives today and the pocket price it would receive if every concession still matched its reason. It is measured against the firm's own ledgers, which already hold every bar. And before anyone shares in it, it has to be measured in a way that both the firm and its customers could look at without flinching, which means counting what was returned to the installer as carefully as what was recovered from the account.
The article this chapter began as ended at the analyst's folder. The book can go further, into the finer points a leader meets when she draws the waterfall in her own firm, and the first of them is which customers to draw.
Not the worst. A waterfall drawn for the customer everyone already complains about produces a story everyone already knew, and a remedy aimed at one account. Draw three: one of the largest accounts, one ordinary account in the middle of the book, and one small, loyal customer who never asks for anything. The three together show the shape of the firm's pricing habits rather than the history of one relationship, and the small customer is the one most likely to show something the firm would not want to defend.
The second point is where each bar lives. The invoice discount is in the billing system. Rebates are usually in a separate file or spreadsheet kept by finance or by the sales director, sometimes in the contract alone. Payment terms are in the customer master, but the days actually taken are in credit control's ageing reports, and the discount for early payment is often applied automatically whether or not the payment was early. Freight is in the carrier's statements, rarely allocated to customers at all. Returns are in the warehouse's log and the credit notes. Promotional and marketing contributions are wherever the person who agreed them keeps them. Before drawing anything, list the ledgers and their owners. The list is itself a finding: it shows how many desks hold a piece of one customer's price.
The third is how to put a cost against a unit. Some bars attach naturally to a product, and some, like freight and the cost of returns, attach to orders or to the year. Spread them in the way the customer actually causes them: freight by the orders that incurred it, not by an average across all customers; returns by the lines returned. A waterfall built on averages hides exactly the differences it exists to show.
The fourth point is a caution about what the waterfall is not. The pocket price is what stays in the firm's pocket from the sale. It is not the profit on the account, because customers differ in what they cost to serve in ways the waterfall does not capture: the time of the account manager, the technical support, the complexity of their orders. A low pocket price can be fair for a customer that is cheap to serve, and a high one unfair for a customer that is cheap to serve and pays it anyway. Read the waterfall beside whatever the firm knows about cost to serve, and say plainly where it knows little.
The fifth is the band, and what to look for in it. Once the pocket prices for one product are sorted, ask what explains the order. Volume should explain some of it. Cost to serve should explain some. Whatever remains, the spread that no reason the firm would defend can account for, is the finding. It usually sorts by account manager, by the year the terms were last negotiated, or by whether the customer has ever threatened to leave.
The sixth is who is in the room. The waterfall is drawn with the account manager, never about him. He supplies what the ledgers cannot: what was asked, what the alternative seemed to be, what the customer values. If he first sees the waterfall in a finance report, the exercise has become an audit of his judgement, and he will defend every bar on it.
And the decision that most changes the result is timing. Draw the waterfall before the customer's annual negotiation, not after it. Every bar is easier to change when the conversation is already due, and almost impossible to change in the middle of a year without seeming to go back on a deal. Chapter 6 places the first waterfalls in the opening weeks of a ninety-day plan, before anything is said to any customer, and Chapter 7 gives the waterfall sheet, the list of ledgers, the three questions for each bar and the band worksheet on single pages.
It is ten to six on a Tuesday, a year later, and the analyst has the same customer's year on her desk. It fits on one desk this time. The rebate was paid in January against volume the customer reached in November. The early-payment discount appears on most invoices and not on the ones paid late. The same-day vans are on the invoices, at a charge the buyer agreed in April, and there are fewer of them, because the customer's stores manager has started ordering a day ahead.
The thermostatic valve's pocket price for this account is a little higher than last year. The installer's is a little lower. The band, when she draws it again, is narrower, and its order makes more sense.
The A3 sheet from last February is in a folder with the customer's name on the spine. She adds this year's beside it, in pencil, and puts the folder back where the next person to look after the account will find it.
Chapter Two
Work delivered and never written down is neither charged for nor received as a gift.
The Favour on a Friday
Haute Lumière · The Price Nobody Charged · Chapter Two
It is twenty to six on a Friday, and the office of a structural engineering practice has mostly emptied for the weekend. An engineer in her sixth year is finishing her timesheet, the last thing she does every week, at a desk by the window where the last of the light has gone flat and grey over the car park. The project is an extension to a primary school: four classrooms and a hall, on a fixed fee agreed in the autumn. The timesheet has a row for the project and a row for "general office", and she is deciding which row this week belongs in.
It was not a planned week. On Monday the architect moved one of the classrooms a metre and a half to save a tree, and the drainage layout had to be redrawn. On Wednesday the client's project manager rang to ask, kindly and apologetically, whether someone could come to site on Thursday, because the contractor had found an old culvert where the drawings showed none. She went. On Thursday evening and Friday morning she answered eleven questions from the contractor that the project manager had forwarded with the line could you just have a quick look at these? None of it was in the scope the fee was written for. All of it was work the project needed.
The fee for the stage she is in was spent two weeks ago. If she books the week to the project, the project report on Monday will show it further over budget, and the director will ask why. If she books it to "general office", the project will look as it looked last week, and nobody will ask. Her shoulders are already up by her ears. She types the hours into the second row.
She is a composite, drawn from the way these Friday evenings are described in the research below and in the house's own practice, and no firm is meant. Anyone who has worked in a firm that sells its time has filled in that row.
Ammanuel Santa Anna's canon has a name for what she has just done, though it was written about families and teams rather than invoices.4 Every relational field, it says, "also contains what family systems theorists have called an invisible ledger — the unspoken accounting of what has been given and received, what is owed and what remains unpaid." The client does not know the week happened. Her director does not know. The only record of it is a row on a timesheet that says, in effect, that it did not.
This chapter makes one claim. A great deal of the revenue a professional firm loses is not lost to clients who refuse to pay. It is work the firm delivered and never wrote down, given away without anyone deciding to give it, and because it was never seen, it was neither charged for nor received as a gift. The cure is not to charge for every favour. It is to make every favour visible, so that the firm chooses, each time, whether it is a sale or a gift.
In 1999 the sociologists Susan Leigh Star and Anselm Strauss published a paper on what they called the ecology of visible and invisible work, in a journal devoted to how people work together with computers.5 Their subject was the way some kinds of labour disappear from the record: the work that is so expected it is never counted, the work done by people whose role makes them hard to see, and the background work that makes the official work possible. Their examples ran from nursing to office systems. Their concern was what happens when organisations design their records, and later their software, around the work they can see, and so lose the work they cannot. The paper is an argument built on years of field studies, not a measurement, and its text was not reopened for this chapter; its central idea has since become a common reference in the study of work.
The engineer's week is invisible work in exactly their sense, and for all three of their reasons. It was expected: a good engineer goes to site when there is a culvert. It was done by a person whose role makes it easy to overlook: a sixth-year engineer, not the director whose name is on the appointment. And it was background: the redrawn drainage does not appear in any deliverable the client will notice, because the drawings were always going to show drainage. The project's records are designed around stages and deliverables. Her week fits neither, so it fits nowhere.
Santa Anna's canon, writing about the roles that hold an organisation together without appearing on its charts, states the consequence in seven words: "You cannot distribute what you cannot see." It was written about recognition and the weight of hidden labour, and it applies with uncomfortable exactness to money. A firm cannot bill what it cannot see. It cannot decide to give away what it does not know it is giving. It cannot even learn that its fees are too small, because the evidence for that has been booked, week after week, to "general office".
The canon is careful not to make the invisibility sound noble. It considers the temptation to praise the quiet people who hold everything together without credit, and answers it bluntly: "No. It is not beautiful. It is often unjust." The engineer will not be thanked for the week, because nobody will know it happened. She will, in the spring, be asked why her utilisation looks low.
The second reason the week was unbilled is older than the week. The fee was too small from the start, and there is good evidence about why fees like it usually are.
In 1994 the psychologists Roger Buehler, Dale Griffin and Michael Ross published a series of studies on what Daniel Kahneman and Amos Tversky had called the planning fallacy.6 They asked people to predict how long their own tasks would take, academic and otherwise, and then checked. The predictions were reliably too optimistic. When people thought aloud while predicting, they focused on how the task would go if it went to plan and paid little attention to how similar tasks had actually gone before. When their past predictions had failed, they explained the failures as the result of particular, passing circumstances, which made the past seem irrelevant to the next estimate. The most useful finding for anyone who writes a fee was in the fourth study. When people were told to connect their prediction to relevant past experience, the optimistic bias disappeared. And when they predicted other people's tasks, they were, if anything, too pessimistic.
These are laboratory and classroom studies, mostly of students, and they measure predictions of time, not the pricing of professional work. But the mechanism they describe is easy to see in the engineer's practice. The fee for the school was written by the director, in the autumn, from a plan: so many drawings, so many site visits, so many meetings. It was not written from the record of the last five school extensions, because that record does not exist in any usable form. It is scattered across timesheets in which the overruns were booked to "general office". The planning fallacy and the invisible week feed each other. The fee is too small because past overruns were hidden, and the overruns are hidden because the fee is too small.
Santa Anna's canon has an observation about project plans that the engineer would recognise: "The relational labor that holds a team together is invisible on the project plan." It is written in the same chapter on keystone roles, and it names the same gap from another angle. The work a fee does not foresee is very often the work of holding things together: the extra call, the site visit, the eleven questions. It is the work clients value most, and the work the plan was least able to see.
A favour is not yet a gift
Haute Lumière · The Price Nobody Charged · Chapter Two
There is a defence of the engineer's Friday, and it deserves to be heard properly, because it is the reason most firms keep doing it.
In 1971 the psychologist Dennis Regan ran a small experiment on favours.7 In the version usually retold, a participant and a confederate rated pictures together; in some sessions the confederate left the room and came back with two soft drinks, one for himself and one for the participant, unasked. Later the confederate asked the participant to buy raffle tickets. Those who had received the drink bought more, and the effect held whether or not they liked him. The study is small, and its details are given here as it is customarily described, not from the paper itself. It has been read ever since as evidence that a favour creates a felt obligation to return it.
A decade later the economist George Akerlof gave the same idea a larger setting.8 Some employers, he argued, pay more than they strictly need to, and workers respond by giving more than the minimum, because effort follows norms about what a fair day's work is, and those norms respond to how the worker is treated. He called the arrangement partial gift exchange. It is a theoretical argument, built on a case of clerks who exceeded their standard, and it has been tested and debated since. Its point for a professional firm is simple. Relationships between firms run partly on gifts, and the gifts are not waste. A client who has been helped generously in a difficult week is more likely to come back, to recommend, and to be reasonable when something goes wrong.
So the director's habit, which the engineer has learned from him, has a real basis. Don't nickel-and-dime a good client is not a foolish rule. It is a rough version of everything Regan and Akerlof found.
But both of them describe a gift the other party knows about. Regan's participant saw the drink arrive. Akerlof's workers knew they were being paid above the market. A favour creates obligation, goodwill and loyalty only when it is received as a favour. The engineer's week was not received as anything. The project manager who asked for the site visit does not know that it fell outside the fee, because nobody told her. As far as she knows, it was part of what she bought. The firm has given away a week and received, in return, a client whose sense of what the fee includes has quietly grown.
That is the precise difference between a gift and a leak. A gift is chosen and named, and it enters the relationship as a gift. A leak is the same work, given without anyone deciding to give it, and it enters the relationship as an expectation.
The canon explains why the leak feels like kindness to the person causing it, and why it so often turns into something else.
Writing about the capacity to ask for what one needs, the canon's chapter on constructive development lists the people who have lost it, and its first example is uncomfortably close to a professional firm: "The therapist who cannot charge a fair fee." It goes on to say what happens to generosity without that capacity: "Without it, generosity becomes martyrdom. Service becomes self-erasure." The engineer does not think of herself as a martyr. She thinks of herself as someone who does not make a fuss. But a firm whose people cannot say that is outside the scope, and here is what it will cost has made a habit of self-erasure, and its accounts show it.
The house's Luminous 100 gives the managerial version a name. In its account of Kim Scott's Radical Candor it describes the quadrant most leaders fall into, high care without challenge, as "Ruinous Empathy (high care/low challenge — most common mistake)". The director cares about the client. He has never challenged the client's sense of what the fee includes, because doing so feels like a failure of care. The result is the opposite of what care intends. The client is never given the chance to decide whether the extra work is worth paying for, and the firm's people carry the cost in evenings.
And the invisible ledger does not stay invisible. The canon is explicit about what happens when it is left unattended:
When one person consistently gives more than they receive, or when sacrifices go unacknowledged, or when one party bears disproportionate cost for shared benefit, the invisible ledger becomes imbalanced. This imbalance generates a kind of relational debt that, if unaddressed, will eventually surface as resentment, withdrawal, or explosive conflict.
In a professional firm the resentment surfaces in familiar places. It surfaces in the engineer, who begins to dread the project manager's calls. It surfaces in the director, who decides at the end of the project that this client is not profitable and quietly stops bidding for its work. And sometimes it surfaces in the client, who is billed at the end for a lump of overrun nobody warned her about, and feels, correctly, that she was never asked.
In a firm that sells time, the hours a team spends that nobody budgets are not only a strain on the people who work them. They are revenue, earned and not recorded, and the record that would show it was emptied, every Friday, into a row called "general office".
The invisible week has a second cost, and it falls on the person who did the most.
In April the practice runs its annual review of utilisation, the share of each engineer's hours booked to projects that clients pay for. It is a reasonable measure, and in a firm that sells time it is one of the few that tell a director whether the practice can pay its people. The engineer's figure is lower than last year's, and lower than two of her colleagues'. The director notices, and asks about it kindly, in the way that makes a person feel worse than a sharp question would.
She knows why the figure is low. It is low because she booked the culvert week, and four other weeks like it, to "general office". Her colleagues, on projects whose fees happened to be larger, booked the same kind of work to the project, because there was still budget to absorb it. The measure did not record who worked hardest or who helped the client most. It recorded whose fees ran out first. The engineer who did the most unpaid work for the practice's clients looks, on the one report the director reads, like the least productive engineer in the room.
This is the cruelty of an invisible ledger inside a firm, and it is worth stating without softening. The people who absorb the overruns are usually the conscientious ones, the people who would rather stay late than tell a client the scope has changed. A system that hides their work does not merely fail to reward them. It records them as the problem. Over a few years, the conscientious learn one of two things. Some learn to stop absorbing, and book everything to the project regardless, which at least makes the overruns visible, though in the worst possible way: at the end, as a surprise. Others leave, for a firm where their work is seen, and take with them the habits of care the practice most needed.
The canon on keystone roles names the purpose of seeing such work, and it is careful about what that purpose is. Its aim, it says, is "to make the invisible visible — and then to advocate for the recognition, compensation, and support that visible contribution deserves." In a professional firm, the first two of those words are plain. Recognition means the director knows the culvert week happened. Compensation means that somebody, either the client or the firm, decides on purpose who pays for it.
None of this makes utilisation a bad measure. It makes it a measure that is only as honest as the timesheets beneath it, and the timesheets are only as honest as the firm's willingness to see an overrun without treating it as a failure. A director who wants accurate timesheets has to make it safe to book a week to a project that is over budget. That is a matter of what he does the first time he sees one. The same is true of a medical practice's front desk, where a question nobody asked becomes a denied claim weeks later (Chapter 4): a record is only as honest as the first person who fills it in, and as safe as the response to what they write.
The ledger on the desk
Haute Lumière · The Price Nobody Charged · Chapter Two
What changes this is less a policy than a piece of paper, kept on the day.
The engineer's director decides to try something on the next project, and the engineer agrees to run it. They call it the scope ledger. It is a single page per project, kept by the person doing the work, and it has one line for every request that falls outside the agreed scope, written on the day the request is made, not at the end of the week. The line says what was asked, by whom, and roughly how long it will take. That is all. Writing it takes a minute.
The page is read every Monday by the project lead and the director together, and every line gets one of three answers, written beside it. Change: the request is real work outside the scope, the client is told the same week what it will cost, and it goes on the next invoice with the client's agreement. Given: the firm decides to do it without charge, on purpose, because the client is a good one or the request is small or the relationship needs it, and it appears on the next invoice as a line at no charge, so that the client sees the gift. Declined: the request is not the firm's to do, and the client is told who it belongs to.
The middle answer is the one that surprises people. It does not reduce what the firm gives. It may increase it. What it changes is that the gift is now a gift. The project manager receives an invoice that says, in one line, site visit following discovery of culvert, Thursday 12th: no charge, and she knows, for the first time, that the firm did something for her it did not have to do. Regan's drink has arrived, and she has seen it.
The client's side of this is worth a moment, because firms tend to assume that clients prefer not to know. The project manager at the school has a budget of her own and a governing body to report to. A firm that tells her, in the week it happens, that a request falls outside the scope and what it would cost gives her something she needs: the chance to decide. Sometimes she will say the extra work is worth it and find the money. Sometimes she will say it can wait, or that the contractor should answer his own questions. Sometimes she will ask the firm to do it as a favour, and the firm will say yes, and both will know a favour has been done. What she will rarely do is thank a firm that tells her at the end of the project, in a single line on a final invoice, that she owes it for a dozen things she never knew she had asked for.
The first answer needs a sentence, and the canon's chapter on governance supplies one that works as well with a client as with a board. It is the sentence it gives a member who holds a commitment the circumstances have outgrown: "I need to request a renegotiation — the scope has expanded beyond what I can deliver in the original timeline." Said early, in a call, before the work is done, it is an ordinary professional conversation. Said at the end, on an invoice, it is a dispute. The difference is not the amount. It is when the client was asked.
The page also does something the director did not expect. By the end of the first project it holds a record of every request that fell outside the fee, with a date and an estimate of time, and when he writes the fee for the next school extension he reads it first. Buehler, Griffin and Ross found that the optimism in an estimate disappears when the person making it is made to look at what happened last time. The scope ledger is the record that makes that possible, and it is the first such record the practice has ever had. The overruns that used to vanish into "general office" are now evidence.
There are predictable ways for it to go wrong. The page becomes a form with twelve fields, and nobody fills it in. The director reads it and answers given to everything, which is the old habit with paperwork attached. The client is told about a change only after the work is done, which turns the conversation back into a dispute. Or the firm begins charging for things it always gave freely, without telling the client that anything has changed, and a relationship built on generosity learns overnight that it was being metered. Each failure is visible on the page itself, which is why the page is read every week by someone with the authority to answer it. Chapter 7 carries the scope ledger as a run sheet, with the page, the three answers, the sentences for the call and the signs that it is failing, and Chapter 6 places it in the order of a first ninety days.
The gain is easiest to see on the timesheet. Every week booked to "general office" that was really project work is revenue the firm earned and did not ask for, or a gift it gave and was not thanked for. Recovering it does not require the firm to become less generous. It requires the firm to know when it is being generous, and to say so. The same discipline matters when the work is not an hour but a promise written into a contract, which is the clause nobody enforced (Chapter 3).
It is also the reverse of the concession that outlived its reason (Chapter 1). One leaks because nobody wrote it down, the other because nobody read it again.
The scope ledger is simple enough to start on Monday and easy to let lapse by the third week. The book can carry the finer points the article left out.
The first is knowing what the scope is. Most out-of-scope work is not recognised as such by the person doing it, because nobody has told her what the fee covers. Before the ledger starts, the project lead reads the appointment or the statement of work with the team and writes, on the first page of the ledger, three or four lines in plain words: what is included, what is not, and the requests that most often arrive and are not included on projects of this kind. In the engineering practice of this chapter the list was short. Changes after the architect's design is frozen. Site visits beyond the number stated. Answering the contractor directly rather than through the client. Most practices find their list is equally short, and that the same three requests recur on every project.
The second is the minute rule. A line is written on the day the request arrives, by the person who receives it, and it takes no more than a minute: the date, who asked, what for, a rough estimate of the time. It does not need the price or the answer. A ledger that asks for more than a minute will be filled in on Fridays from memory, and memory is exactly where the invisible week went.
The third is who answers. The three answers, change, given and declined, are given by someone with the authority to commit the firm, usually the project lead with the director, once a week. The person who wrote the line should never have to decide alone whether to charge for her own work. That decision is what she was avoiding on Friday evening, and the ledger works because it takes it out of her hands.
The fourth is the sentence for the client. A change is raised in the week it arises, by telephone or in the next meeting, before the work is done where that is possible, and confirmed in writing in two lines. The words matter less than the timing, but a plain form helps: what was asked, that it falls outside what was agreed, what it will take, and a question, would you like us to go ahead? Where the work has to be done at once, as the culvert visit did, the call is made the same day and the confirmation follows.
The fifth is the named gift. A request answered given appears on the next invoice as a line with its description and no charge. It is not hidden in a discount or left off. Some firms are uncomfortable with this at first, because it seems to draw attention to generosity. It does. That is the point. A gift the client has seen is a gift; a gift the client has not seen is a leak that the client experiences, later, as the normal level of service.
The sixth concerns fixed fees and time-based fees, because the ledger works differently in each. On a fixed fee, the ledger protects the fee: it is the record from which changes are priced and the evidence for the next fee. On a time-based fee, the danger is the opposite. Out-of-scope hours are billed without the client having been asked, and the client discovers them at the end. The ledger serves both by the same rule: the client hears about the request in the week it happens.
And the decision that most changes the result is what the director does with the ledger when he writes the next fee. Read it first. Count the lines on the last few comparable projects, and what they took. That reading is the outside view the planning-fallacy research found removes optimism from an estimate, and it is the first time most practices have had one. A second change is worth making at the same time: the utilisation report shows absorbed hours, the hours booked to a project over budget, as their own column, so that the engineer who absorbs the most is seen as the engineer who did the most. Chapter 6 places the ledger in the ninety days, after the waterfall and before the register, and Chapter 7 gives the ledger page, the scope list, the sentences for the call and the weekly review.
It is twenty to six on a Friday in June, and the engineer is finishing her timesheet at the same desk by the window. The light is longer now. The project is a medical centre, and its scope ledger has nine lines on it, each with an answer beside it. Four are changes, agreed with the client by phone and on the next invoice. Three are gifts, and the client's project manager mentioned one of them in a meeting on Tuesday, unprompted, with thanks. Two were declined, politely, and passed to the architect.
The timesheet still has a row for "general office". She does not use it this week.
She books the week to the project, saves it, and goes home.
Chapter Three
An unused clause is a promise to oneself that broke, and the remedy is a decision, taken in the open.
The Clause in the Drawer
Haute Lumière · The Price Nobody Charged · Chapter Three
It is eight o'clock on a Wednesday in March, and the commercial manager of a firm that supplies cleaning chemicals to food factories has a contract open on her desk. It is a five-year supply agreement with the firm's second-largest customer, a maker of chilled ready meals, signed four years ago. The new finance director asked last week for a list of every contract with a price-indexation clause, and this is the fourth she has read this morning. Her coffee is going cold in a mug with the customer's logo on it, a gift from their last site visit.
Clause nine is the one she was looking for. On each anniversary of the agreement, the prices in the schedule move with a published index of chemical costs, up or down. The anniversary is the first of March. It has passed three times. The prices on the invoices are the prices in the original schedule. The index has risen every year.
There is more, now that she is reading slowly. Clause twelve sets a minimum annual volume, with a charge for any shortfall; the customer fell short in the second year when it closed a line, and no charge was raised. Clause fifteen sets payment at thirty days, with interest on late payment; the customer pays, on average, at seventy-odd. She has known the customer's head of procurement for nine years. They were at a trade dinner together in January. She feels it in her chest, the small drop of realising that a conversation she has avoided without knowing it has just become unavoidable.
She is a composite, drawn from the way these mornings are described in the research below and in the house's own practice, and no firm is meant.
Ammanuel Santa Anna's High Frequency Workbook is written about a person's relationship with money, and it lists, among the reasons people undercharge, a sentence that describes the commercial manager's chest exactly:9 "Raising your prices feels like hurting someone. Negotiating a salary feels like being aggressive." She does not think of clause nine as a price rise. She thinks of it as something she would have to do to someone she likes.
This chapter makes one claim. The terms a firm has already agreed, and does not apply, are among the largest sources of revenue it has earned and not collected, and they go unapplied less from carelessness than from a well-founded fear of what applying them will do to a relationship. The answer is not to enforce every clause. It is to make every unused clause a decision, taken in the open, at the time it falls due.
The best-known study of how businesses actually use contracts was published in 1963 by a law professor at the University of Wisconsin, Stewart Macaulay.10 He wanted to know what contracts did in practice, so he asked. His method, as he described it, "involved interviewing 68 businessmen and lawyers representing 43 companies and six law firms," and all but two of the companies had plants in Wisconsin. His preliminary finding is in the first lines of the paper: "businessmen often fail to plan exchange relationships completely, and seldom use legal sanctions to adjust these relationships or to settle disputes."
The detail is what has kept the paper in print for sixty years. Macaulay describes the battle of the forms, in which a buyer's purchase order and a seller's acknowledgment carry different and inconsistent terms on the back, both are filed, and both sides believe they have a contract. A manufacturer of packaging audited its own orders for five days in four different years to see how often it had failed to agree terms with its customers; in 1953 the figure was 75.0 per cent of orders, and in 1956 it was 59.5. Purchasing agents expected to cancel orders freely; sales managers accepted that they would be cancelled. Lawsuits were rare. Macaulay also noticed where the care in a contract tended to go: "It is likely that businessmen pay more attention to describing the performances in an exchange than to planning for contingencies or defective performances or to obtaining legal enforceability of their contracts." The product and the price are specified closely. What happens when something changes is left to the telephone. And the reason was not ignorance of the law. It was the relationship. One purchasing agent, in a passage Macaulay quotes at length, put the attitude that ran through his interviews: "You don't read legalistic contract clauses at each other if you ever want to do business again. One doesn't run to lawyers if he wants to stay in business because one must behave decently." Another businessman put it more briefly: "You can settle any dispute if you keep the lawyers and accountants out of it. They just do not understand the give-and-take needed in business."
It is a study of one state, in one decade, built on interviews, and Macaulay called it preliminary. Its strength is that it reports what people did, in their own words, rather than what the law assumed they did. Read now, it is less a description of 1960s Wisconsin than of a disposition that most commercial managers will recognise in themselves. The contract is written to be put in a drawer. The relationship is conducted by telephone. Clause nine is, in Macaulay's sense, a legalistic clause, and the commercial manager has not read it at the customer for three years, because she wants to do business again.
What Macaulay saw as decency has a cost he did not set out to count. The give-and-take he admired is real, and it keeps relationships alive. But give-and-take implies both directions. A clause that is never applied, by a supplier that never asks, is not give-and-take. It is give.
The commercial manager's reluctance has a second source, and it is measurable. Changing a price, even one the contract allows, is work.
In 2004 a team of five researchers, Mark Zbaracki, Mark Ritson, Daniel Levy, Shantanu Dutta and Mark Bergen, published a study of exactly that work at a large American industrial manufacturer and its customers.11 Economists had long theorised about the cost of changing prices, usually as the physical cost of printing new price lists, which they called menu costs. The team measured it directly. They found that the cost of adjusting prices was "a much more complex construct" than the literature had recognised. Beyond the menu costs, they identified managerial costs (gathering information, making the decision and communicating it) and customer costs (communicating with customers and negotiating with them). Their abstract gives the proportions plainly: "the managerial costs are more than 6 times, and customer costs are more than 20 times, the menu costs." In total, at this firm, price adjustment cost 1.22 per cent of revenue and a little over a fifth of its net margin. And the study closed on a finding that every commercial manager knows from the inside: the team documented "managers' fear of antagonizing customers."
It is one firm, studied in depth, and the proportions belong to it. Their significance is not the exact figure. It is the shape. Most of the cost of changing a price is not in the system. It is in the conversation with the customer, and the managers who must hold that conversation feel its cost in advance, which is why they put it off. A firm that wonders why its indexation clauses go unused has its answer in that study. Applying the clause is cheap. Telling the customer is expensive, and the person who must tell them is rarely measured on whether they did.
The fear, it turns out, is not irrational. In 2010 the economists Eric Anderson and Duncan Simester published the results of a field experiment run over twenty-eight months with more than fifty thousand customers of a retailer.12 They wanted to know what happened when customers bought a product and later saw the same retailer sell it for less. The customers bought less from the firm afterwards. The effect was largest among the firm's most valuable customers, those who had bought most recently and at the highest prices. It is a study of retail customers and of falling prices rather than rising ones, and its setting is a long way from a supply agreement. But it shows something important for the commercial manager. Customers keep their own ledger of how a supplier has treated them on price, and they act on it, and the ones who act most are the ones a firm would least like to lose.
So the drawer has two reasons to stay shut. The relationship, which Macaulay's businessmen were right to protect. And the cost and risk of the conversation, which Zbaracki and Anderson measured. Neither is a reason to leave clause nine unread. Both are reasons to read it differently.
Three anniversaries, nobody's diary
Haute Lumière · The Price Nobody Charged · Chapter Three
It is worth tracing how a clause like clause nine actually goes unused, because it is almost never a decision.
The agreement was negotiated by the commercial manager's predecessor and signed by the managing director. It was scanned by the legal team and filed in the contracts folder. The price schedule was keyed into the billing system by an administrator, who set the prices and had no field in which to record that they were due to move each March. The first of March is in nobody's diary. The finance team, which sees the margin on the account fall a little each year as chemical costs rise, reads it as a market problem. The account manager, who visits the customer every month, does not know clause nine exists, because nobody gave him the contract; he was given the price list. The commercial manager knew in the abstract that the agreement had indexation, and assumed that billing applied it.
Each of these people did their job. The clause fell into the space between the desk that signed it, the desk that filed it and the desk that bills from it. It is the same space the discount below the invoice (Chapter 1) falls into, and the same space a professional firm's unbilled week falls into: the place where a promise the firm made to itself is held by nobody.
Santa Anna's canon on governance begins its account of broken promises with three short sentences that fit a supply agreement as well as a board: "Promises break. This is not a possibility to be guarded against. It is a certainty to be prepared for." A contract is a set of promises made in both directions. Some of the customer's promises, the volume and the payment terms, have broken. Some of the supplier's promises to itself, to apply the index on the first of March, have broken too. Neither side has said so. The canon's point is that the breaking is not the failure. The failure is the silence afterwards.
The canon goes on to distinguish two ways of meeting a broken promise. "Conventional accountability treats broken promises as failures requiring consequences." The other way treats them as ruptures to be repaired. For the commercial manager, conventional accountability would mean a letter, three years of indexation back-billed, a shortfall charge and an interest claim, sent by the finance director. It would be legally sound. It would also, by every finding in this chapter, cost the firm its second-largest customer.
The commercial manager's difficulty has a shape the house's Luminous 100 names well. In its account of Barry Johnson's polarity management, the entry begins: "Some challenges aren't problems to solve but polarities to manage". Enforcing contracts and keeping relationships is one of them. A firm that enforces every clause, every time, gets the revenue and loses the give-and-take Macaulay's businessmen depended on. A firm that enforces nothing keeps the warmth and loses the revenue, and in time loses the warmth as well, because the supplier who never asks begins, privately, to resent the customer who never offers.
The canon's chapter on conflict describes what holding only the relationship pole looks like from the inside, in a passage that reads like a commercial manager's year: "We smile when we are angry. We agree when we disagree. We swallow truths that need to be spoken and speak pleasantries that need to be swallowed." She was pleasant at the trade dinner in January. She did not mention clause nine. She did not, at the time, know she was not mentioning it, which is the most expensive kind of silence there is.
Getting the benefit of both poles is not a compromise between them. It means applying the contract in the way the relationship deserves: early, openly, specifically, and with room to choose. The canon's account of repair gives the choice its exact form. "Repair is not complete until the governance body moves forward — either with a revised commitment that addresses the original breach, or with a conscious decision to release the commitment if circumstances have made it no longer viable." Read at a supplier's desk, that sentence offers three honest answers to an unused clause, not one. Apply it. Revise it, together. Or release it, consciously, and say so.
Before the commercial manager picks up the phone, it is worth sitting for a moment on the other side of the table, because the conversation she fears is not the conversation the buyer will have.
The head of procurement at the ready-meals firm has a manager too. She has a budget for cleaning chemicals, agreed in the autumn, and a finance director who asks every month why the cost of goods has moved. She negotiated the supply agreement four years ago and knows, in the abstract, that it has an index in it. She has assumed for three years that the supplier would apply it, and has budgeted, some years, for a rise that never came. The unspent money went elsewhere. She did not raise it, because a buyer is not paid to ask suppliers to charge more.
What she fears is not the index. It is a surprise. Three years of indexation presented at once, as a back-bill, would land in the middle of her budget with no warning and no explanation she could give her own finance director other than that she had not been watching. A shortfall charge for the year the line closed would arrive two years after the closure was announced, which would make her look as if she had hidden something she had in fact discussed with the supplier at the time. The commercial manager imagines the buyer as the person who will be angry. The buyer, if the supplier gets this wrong, will be the person who is embarrassed.
This changes the conversation. A clause applied every March, on a date the buyer can put in her diary, at an index she can look up herself, is the easiest kind of price change there is to budget for. It is the clause applied once every four years, by surprise, that is hard. Zbaracki's finding, that the customer side of a price change is the most costly part of it, is about this. The cost is not in the change. It is in the negotiation that a surprise forces, and a change both sides could see coming needs very little negotiation at all.
The High Frequency Workbook says of money conversations that they are rarely about money alone: "Money work is emotional work. It is somatic work." The commercial manager's chest and the buyer's budget are both part of the call she is about to make. Knowing that the buyer has her own reasons to want the clause applied properly is what lets the commercial manager make the call as a colleague rather than a creditor.
The register and the call
Haute Lumière · The Price Nobody Charged · Chapter Three
What the commercial manager does next is the practice this chapter is about.
She starts a register. It is a single sheet for each live contract, and it lists only the clauses that move money: indexation and price review, minimum volumes and shortfall charges, payment terms and interest, rates for changes and extra work, rebate conditions, the dates of renewal and expiry. For each clause it records three things: the date it next falls due, the person who will act on it, and what was decided last time. The finance director agrees that the register, not the contracts folder, is the place a clause lives once it is signed. The first full pass, across the firm's forty largest contracts, takes her and an administrator most of a fortnight. It is the first time anyone has read them all.
Then, for each unused clause, they take one of three decisions, and write it down. Apply: the clause is sound and the firm will use it from the next date it falls due, after telling the customer. Revise: the clause no longer fits, because the index is the wrong one, or the volume was set for a business the customer no longer runs, and the firm will propose a change. Release: the firm decides, on purpose, not to use the clause, for a reason it can state, and tells the customer it has done so. The third decision is the one that surprises people, as the named gift does in a professional firm's scope ledger (Chapter 2). It does not cost the firm more than silence does. It costs exactly the same, and it turns a leak into a gesture the customer can see.
For the ready-meals customer the decisions are these. The indexation will apply from this March, at the index's current level, without recovering the three years missed; those years are released, in writing, as the firm's own oversight. The volume minimum will be revised, because the customer closed a line and the old figure no longer describes its business. The payment terms will apply, with the interest clause released for the past and kept for the future.
She rings the head of procurement before anything is sent. The conversation has an order, and the order is most of the skill. She names the clause and the date, and says plainly that the firm did not apply it and that the missed years are the firm's to absorb, not the customer's. She says what will happen from now, and why. She asks about the line that closed, and proposes a new minimum that fits the business as it is. And she asks about the payments, as a question, not an accusation. It turns out that the customer's payment run has been set to seventy-five days for every supplier for two years, by a finance team that never told procurement. The head of procurement had not known either.
The call takes forty minutes. It is uncomfortable for the first five and ordinary after that. The canon on acknowledgment insists that agreements are kept alive by being attended to, that they are "not pro forma rules posted on a flip chart and forgotten." The register is the firm's way of attending to them, and the call is what attending sounds like. Zbaracki's customers were right that a price conversation is costly. What the commercial manager discovers is that most of its cost was in the three years of not having it.
There are predictable ways for this to go wrong. The register becomes a legal exercise, and the first a customer hears of it is a letter. Past years are back-billed by default, which recovers money once and loses the account. The account manager, who holds the relationship, is left out, and learns about clause nine from the customer. Or the register is built once and never read again, and the next anniversary passes like the last three. Each failure has a cure in the order of the work: the register first, then the decision, then the call, and only then the invoice. Chapter 7 carries the register and the call as a run sheet, with the clauses to look for, the three decisions, the order of the conversation and the signs that it is failing, and Chapter 6 places it in the order of a first ninety days.
The gain is plain once the register exists. It is the revenue the firm's own contracts already entitle it to, from the next date each clause falls due, less whatever the firm decides, openly, to release. It is measured against invoices the firm already raises. And because the missed years are released rather than reclaimed, it is a gain the customer can live with, which is the only kind worth measuring. How such a gain is counted honestly, and why the person who negotiates a price should not be paid in a way that teaches them to give it away (Chapter 5), is a question of incentives and measurement as much as of contracts.
The register takes a fortnight to build and a few hours a month to keep. The book can carry the finer points of both.
The first is where the contracts are. In most firms the signed agreements are in a legal folder, but they are not the only place terms live. Macaulay's battle of the forms is still fought every day: the terms on the back of a customer's purchase order, the terms on the firm's own acknowledgement, the price schedules attached to emails, the side letters agreed by a previous sales director. Start with the forty largest customers by revenue and, for each, gather everything that could set a price or a payment: the agreement, its schedules and amendments, the standard terms each side claims, and any letter that changed them. Where two documents disagree, note it; the disagreement is itself a term nobody has settled.
The second is what to read for. Only the clauses that move money go into the register. Price review and indexation, with the index named and the date. Minimum volumes, shortfall charges and take-or-pay terms. Payment days, discounts for early payment and interest on late payment. Rates for changes, extra work and urgent orders. Rebate conditions and their thresholds. Charges for delivery, packaging, storage or returns that the contract allows. Renewal, notice and expiry dates. Everything else in the contract matters, but it belongs to the lawyers, and a register that tries to hold all of it will hold nothing for long.
The third is the calendar. Each clause in the register has a next date and a named person, and the date goes into that person's diary with a month's notice. That one step, putting the first of March in somebody's diary, would have saved the commercial manager three years. It is so simple that most firms assume it has already been done.
The fourth is the three decisions, and the reasons that justify each. Apply, when the clause still describes the business and the customer was told when the agreement was signed that it would be used. Revise, when the clause no longer fits: the index has drifted from the firm's real costs, the volume was set for a business the customer no longer runs, the payment days are out of line with the market both firms work in. Release, when the firm decides that the relationship, the history or its own oversight makes the clause wrong to use for a period. A release is written down with its reason and its period, and the customer is told. A release without a period becomes the old silence with a signature on it.
The fifth is the past. The question every commercial manager asks first is whether to recover what was missed. The default this book recommends is to release it, openly, as the firm's own oversight, and to apply the clause from the next date it falls due. There are exceptions, and they are decisions for the firm's leadership and its advisers rather than for a register; nothing here is legal advice. But a firm that reclaims years of missed indexation from a customer who budgeted in good faith on the invoices it received will usually recover the money once and the relationship never.
The sixth is the order of the call, which is most of the skill. The person who holds the relationship makes it, with the commercial manager if she is not that person. Name the clause and the date. Say what the firm did not do, and that the missed period is the firm's to absorb. Say what will happen from now, and why. Ask about the customer's side, because there is nearly always one, as the seventy-five-day payment run showed. Confirm it in writing, in a few lines, the same week, and only then change the invoice.
And the decision that most changes the result is keeping the register alive. It is read once a month by the commercial manager and the finance director together, in a half-hour meeting with one agenda item: what falls due in the next sixty days, and who is calling whom. Chapter 6 places the register in the middle weeks of the ninety days, and Chapter 7 gives the register page, the list of clauses to read for, the three decisions with their criteria and the order of the call.
It is eight o'clock on a Wednesday the following March, and the commercial manager has the register open, not the contracts. It shows the ready-meals customer's indexation due on the first, at the index published in February, and a note beside it that the head of procurement was told in January, by phone, what it would be. The revised minimum has been met. The payments arrive at thirty-four days on average, since the customer's finance team reset its run.
There are nine other dates in the register this month, each with a name beside it. Two are marked release, with a sentence explaining why.
The mug with the customer's logo is on the desk, and the coffee in it is still hot.
Chapter Four
A denied claim is usually born at the first desk, weeks before it comes back.
The Claim That Came Back
Haute Lumière · The Price Nobody Charged · Chapter Four
It is a quarter past seven on a Wednesday, and the billing coordinator of a family medicine practice is at her desk before the first patients arrive. The practice has eight doctors and serves a mixed neighbourhood: families with employer insurance, older patients on Medicare, a good number on Medicaid. Her screen shows the week's denials, claims the insurers have sent back unpaid. There are a hundred and forty of them. She has time, realistically, to work about sixty before Friday.
The first one is ordinary. A woman was seen in January for a chest infection, and the claim was sent to the insurer on her file. The insurer has replied that her coverage ended on the thirty-first of December. She changed jobs over the holidays. Her new card was in her purse the whole time. Nobody at the front desk asked to see it, because she had been a patient for six years and her details were on the system.
The coordinator knows what happens next. She will ring the patient, get the new details, and resubmit the claim to the right insurer, and if she is quick it will be paid. If she is not quick, it will pass the new insurer's deadline for filing, and it will be written off. Some of the hundred and forty will be. Her jaw is set in the way it is set every Wednesday, before she has opened the second claim.
She is a composite, drawn from the way these mornings are described in the research below and in the house's own practice, and no practice is meant. The research, as it happens, is unusually precise about her week.
Ammanuel Santa Anna's canon, writing about the cost to people of work that has lost its connection to purpose, uses a healthcare example that hangs over this whole subject.13 It describes "the slow erosion that happens when a healthcare worker is forced to prioritize billing metrics over patient care." Anyone who writes about medical billing has to keep that sentence in view. The argument of this chapter is not that doctors should think more about money. It is that a practice which bills well, at the first attempt, frees its people to think about money less.
This chapter makes one claim. A denied claim is revenue earned for care already given, and in most practices it is lost not at the desk that receives the denial but at the desk where its cause began, weeks earlier, in a question nobody asked. The leak is fixed upstream, at the first desk, and the practice that fixes it gains the money, the time spent chasing it, and, as the research shows, the patients it would otherwise have learned to avoid.
Most of what is said about denials comes from consultancies and software vendors, and its figures are rarely traceable to a study. The academic work is smaller, more recent and better grounded, and it begins with the paperwork itself.
In 2018 the economists Joshua Gottlieb, Adam Hale Shapiro and Abe Dunn published a paper in Health Affairs on the complexity of billing for physician care.14 They built new measures of how hard it is to get paid, using a large sample of what the industry calls remittance data, the detailed records insurers send back about each claim, for 2013 to 2015. They found, in their own words, "dramatic variation across different types of insurance." Fee-for-service Medicaid was the hardest to bill, with a denial rate 17.8 percentage points higher than fee-for-service Medicare. On what they called conservative assumptions, they estimated that the health care sector deals with eleven billion dollars a year in challenged revenue, and that the figure could be as high as fifty-four billion.
Five years later three of the same authors, with Matthew Sonnenstuhl and Pietro Tebaldi, published a fuller study in the Quarterly Journal of Economics under a title that puts the subject plainly: A Denial a Day Keeps the Doctor Away.15 They followed the back-and-forth between a large sample of American physicians and many different insurers, claim by claim, through denials and resubmissions, and used it to estimate what billing problems cost doctors: the cost of the haggling, plus the revenue never collected at all. Their estimate is that "physicians lose 18% of Medicaid revenue to billing problems, compared with 4.7% for Medicare and 2.4% for commercial insurers."
That is the leak, measured. It is not the price an insurer pays for a visit. It is the share of that price that never arrives, or arrives only after work that costs almost as much as it recovers. And it differs enormously by who is paying, which is the first clue that it is not caused by the doctors.
The variation is worth dwelling on, because it says where the leak lives. The same doctor, seeing two patients with the same complaint on the same afternoon, writing the same kind of note, can be paid at the first attempt for one and refused for the other, because the two patients' insurers have different rules, different forms and different ways of saying no. The work of medicine is the same. The work of being paid for it is not. That is why billing problems cannot be fixed by asking doctors to be more careful, and why they are so often treated as nobody's problem in particular. They live in the interface between a practice and dozens of payers, each with its own requirements, and a practice that does not know which of its own desks meets which requirement has no way of seeing where its revenue goes.
The coordinator's hundred and forty denials sit on top of a cost that exists even when every claim is paid first time. In 2018 Phillip Tseng, Robert Kaplan, Barak Richman, Mahek Shah and Kevin Schulman published a study in JAMA of the administrative cost of physician billing at a large academic health system.16 They used a method borrowed from management accounting, time-driven activity-based costing, interviewing administrators and physicians to map the path a claim takes through what the industry calls the revenue cycle, and then costing each step.
Their estimates for a primary care visit were thirteen minutes of billing and insurance-related work and about twenty dollars and fifty cents. For a discharged emergency visit, thirty-two minutes and about sixty-one dollars. Measured against the professional revenue of each encounter, they estimated that billing costs ran to 14.5 per cent for primary care visits and 25.2 per cent for emergency visits. It is one health system, studied by interview rather than by stopwatch, and its figures are estimates. But it gives the coordinator's week a scale. Every claim already costs the practice real time to send. Every denial costs that time again, and more, to recover.
This is the reason the house is interested in medical billing at all, and it is worth saying plainly what that interest is and is not. It is not an interest in charging patients more, or in coding visits more aggressively than the care given warrants. Both are wrong, and the second is fraud. It is an interest in the revenue a practice has already earned, for care it has already given, at prices it has already agreed with insurers, which it then fails to collect because of something that happened at its own front desk. That revenue belongs to the practice by every standard anyone would apply. Its loss is pure waste.
The patients a denial costs
Haute Lumière · The Price Nobody Charged · Chapter Four
The most important finding in the Quarterly Journal study is not about money. Dunn and his colleagues used physicians who moved between states, and practices whose patients live on both sides of a state line, to ask what doctors do when billing a particular insurer is hard. They found that physicians respond to billing problems "by refusing to accept Medicaid patients in states with more severe billing hurdles." And they found that these hurdles were, in their words, "quantitatively just as important as payment rates for explaining variation in physicians' willingness to treat Medicaid patients."
Read slowly, that is a finding about who bears the cost of a denial. It is not only the practice. It is the next patient on Medicaid who rings for an appointment and is told the practice is not taking new patients with her insurance, because the practice has learned, one denied claim at a time, that treating her is unusually hard to be paid for. The authors' own conclusion is that administrative frictions "have first-order costs for doctors, patients, and equality of access to health care."
That changes what fixing a denial is for. A practice that bills its Medicaid claims correctly at the first attempt is not only recovering its revenue. It is removing one of the reasons it might otherwise close its doors to the patients who most need them. The gain from fixing the leak does not come at anyone's expense. It is net generative in the plainest sense: the practice is paid what it was owed, the insurer pays what it agreed, and the patient can be seen.
The coordinator's first claim of the morning shows how a denial travels, and it is worth following it backwards.
It was born in January, at the front desk, in the thirty seconds between the patient giving her name and being asked to take a seat. The receptionist confirmed her date of birth and her address. She did not ask whether the patient's insurance had changed, because the practice's routine was to check insurance cards at the first visit and when a patient mentioned a change. The claim was created that afternoon from the visit record, coded by the doctor's notes, and sent automatically overnight. It travelled to the old insurer, which took three weeks to reply. The reply arrived in the coordinator's queue in the middle of February, sorted by the value of the claim, which was small. She reached it in March.
Every person in that chain did their job as it was defined. And the one fact that would have prevented the denial, a new insurance card, was available at the front desk in January, in the patient's purse, for the cost of one question.
The pattern repeats across the queue, with different desks. A denial for a missing prior authorisation began in the referral, when nobody checked whether the insurer required approval before an MRI. A denial for a code the insurer did not accept began in the doctor's note, which did not record the detail the code needed. A denial for late filing began in the coordinator's own queue, where a small claim waited too long behind larger ones. In each case the denial is discovered by the person with the least power over its cause, weeks after the cause, in a department that does not speak to the desk where it began.
Santa Anna's canon on keystone roles describes the work that sits in exactly that gap: "The translational work that bridges organizational silos is invisible on the org chart." The coordinator does translational work all day. She translates insurers' refusals into phone calls to patients and messages to doctors. She does not translate them back into a change at the front desk, because nobody has asked her to, and because the front desk does not report to her.
Readers of Where the Flow Stops will recognise her from its chapter on the night nurse who fetched clean linen from another ward. The researchers who observed that nurse called her kind of response first-order problem solving17: the immediate problem is fixed, and the cause is not reported to anyone who could remove it. The coordinator resubmits sixty claims a week. Each resubmission is first-order. The front desk's question is never changed, and next January the same denial will arrive.
There is a reason the practice does not see this, and it is in how the losses are recorded. A claim that is written off after the filing deadline passes appears in most practice systems as an adjustment, filed alongside the ordinary differences between what a practice bills and what its contracts with insurers pay. Nobody looking at the practice's accounts can tell a lost January chest infection from a routine contractual discount. The canon describes the norms of a group as "shared, often unspoken agreements about what is expected, permitted, and prohibited within the relational space." The practice's unspoken agreement is that denials are billing's problem. Its accounts are built to agree.
Two desks in the chain deserve their own account, because they are the ones a practice is most tempted to treat badly.
The first is the doctor's. A share of the coordinator's queue every week is documentation: claims the insurer refused because the note did not record what the code required, the duration of a visit, the detail of an examination, the reason a test was ordered. The temptation is to send these back to the doctors as complaints, one at a time, with a request to amend the note. That is the route the canon's sentence about billing metrics warns against. Tseng and his colleagues estimated that the physician's own share of billing work for a primary care visit was small, a median of about three minutes, against thirteen for the visit as a whole. Those three minutes are taken from the end of a clinic, from the time a doctor has to think about the patient she has just seen. A practice that adds to them, claim by claim, is buying its revenue with the attention it exists to give.
The better route is the one the walk takes. Documentation denials are sorted by the kind of detail that was missing, and the pattern is shown to the doctors once a month, in ten minutes at a meeting they already hold: these four things, missed this often, cost this many claims. The doctors decide, together, how their notes will capture them, often with a change to the template the notes are written in rather than to anyone's habits. A pattern shown is information. A claim returned is an interruption. The difference is whether the doctor experiences billing as a system she helps design or as a stream of small accusations arriving after her clinic.
The second desk is the patient's kitchen table. A claim that is denied, and not resubmitted in time, does not always disappear into the practice's adjustments. Sometimes it becomes a bill, sent to the patient, for care she believed was covered. The woman with the chest infection did nothing wrong; she changed jobs over the holidays, and her new insurance would have paid. If the coordinator does not reach her claim before the deadline, the woman may receive a letter in the spring asking her to pay for a visit she had forgotten about, for reasons nobody explains. She will ring the practice, upset, and the person who answers will be the receptionist who did not ask about her card.
That letter is the leak arriving at the one person in the chain who had no part in causing it. It costs the practice the revenue, and then it costs the practice the patient's trust, and it costs the receptionist a difficult phone call she will not connect, unless someone shows her, with the question she was never told to ask. Recovering the claim at the first attempt is the only version of this story in which nobody is hurt.
The denial walk
Haute Lumière · The Price Nobody Charged · Chapter Four
What changes this is a way of reading the queue, and it begins by sorting it differently.
The coordinator asks the practice manager to give her two hours on a Thursday, away from the queue, with the last month's denials printed out. They sort them, not by value, as the queue does, but by cause: coverage and eligibility, prior authorisation, documentation and coding, filing deadlines, everything else. The sort takes most of the first hour. When it is done, one pile is plainly larger than the others. More than a third of the month's denials began with insurance details that had changed since the patient's last visit.
Then they walk three claims from that pile backwards, as the coordinator's first claim was walked above, desk by desk, from the denial to the moment the patient arrived. At each desk they ask what the desk knew, what it did not know, and what single question would have closed the gap. All three walks end at the front desk, at the same thirty seconds.
The change they make is small, and it is made at the first desk. From the next Monday, every patient is asked at every visit, Has your insurance changed since we last saw you?, and the receptionist scans the card if it has. The practice manager explains it to the front desk as what it is, a way of making sure the practice is paid for the care the patient is about to receive, and so of keeping the practice able to see patients like her. The receptionists, it turns out, had noticed for years that some patients mentioned new jobs in passing. Nobody had told them it mattered.
The first month is not smooth. Some patients are irritated to be asked the same question at every visit, and the receptionists learn to explain it in a sentence: the practice wants the right insurer to pay, so that nobody gets a bill by mistake. One receptionist asks whether she will be blamed for the denials that still come through, and the practice manager answers, in front of the others, that the walk is about the question, not about who forgot to ask it. By the fourth week the receptionists have started catching other things: a patient whose employer has changed insurers without telling her, a child added to a parent's plan under a new member number. They write them on a sheet by the desk, and the coordinator collects it on Fridays. Without anyone designing it, the front desk has become the first place billing problems are found rather than the first place they are made.
The house's Luminous 100 has a sentence about the most demanding process discipline in modern operations that fits this exactly. In its account of Six Sigma it advises: "Apply DMAIC's rigor to processes that genuinely benefit from standardization". The front desk's thirty seconds is such a process. The same entry is careful to separate the variation worth removing from the variation worth protecting, and it distinguishes "between wasteful variation (genuine inefficiency to reduce) and generative variation (developmental experimentation to protect and nurture)". A receptionist's warmth with a frightened patient is the second kind, and nothing in the new question touches it. A missed insurance card is the first.
Two further habits make the walk stick. The denials are counted every week by cause and written on the board in the staff room, so that the front desk sees its pile shrink and the doctors see theirs. And every write-off is recorded with its reason, not as an adjustment, so that the practice can read, for the first time, what it is losing and where. The canon on acknowledgment describes the right expectation for practices like these: an "ongoing organizational practice — more like hygiene than surgery." The walk is not done once. It is done every month, on whichever pile is largest, and the largest pile changes.
There are predictable ways for it to go wrong. The walk is done on the largest claims rather than the most common causes, and fixes the rare expensive problem while the ordinary one continues. The new question becomes a form the patient fills in and nobody reads. The doctors' pile, documentation, is handed to the doctors as a complaint rather than shown to them as a pattern, and they hear it, as the canon warns, as one more demand to put billing ahead of care. Or the practice, pleased with its recovered revenue, begins to code more aggressively, which is not recovery at all. Chapter 7 carries the denial walk as a run sheet, with the sort, the three questions at each desk, the board and the signs that it is failing, and a version of it for firms outside medicine; Chapter 6 places it in the order of a first ninety days.
The practice is not unusual, and the walk is not specific to medicine. A manufacturer whose customers deduct money from its invoices, for short deliveries, late arrivals, damaged pallets or promotions they claim were agreed, has a denial queue of its own, worked by someone in credit control who resubmits and writes off and never tells the warehouse. The same sort by cause, the same walk to the first desk, applies. So does the lesson of the unbilled week in a professional firm (Chapter 2): revenue that is earned and not recorded at the moment it is earned is revenue that the organisation will spend money, later, failing to recover.
And the gain, before anyone shares in it, has to be measured against the practice's own history, in a form both sides could sign (Chapter 5), which for a practice means counting denials prevented, not only denials won.
The denial walk is the most specific practice in this book, because the setting is specific, and also the one that travels furthest. The book can carry the finer points of both.
The first is the sort. Use the last full month of denials or, outside medicine, of short payments and customer deductions. Sort them by cause, in a few plain categories the team can apply without argument: who is covering the patient or the account; permission that was needed in advance and not obtained; what the record said or failed to say; deadlines missed; and everything else. The insurers' and customers' own reason codes help, but they are written for their purposes, not yours, and several codes often describe one cause. Sort by what would have prevented the denial, not by what the payer called it.
The second is choosing the claims to walk. Take three from the largest pile, ordinary ones, of ordinary value. A single large claim is tempting and misleading, because large claims are often unusual. Walk each backwards from the denial to the moment the patient arrived or the order was placed, and at every desk ask what it knew, what it did not know, and what single question would have closed the gap. Write the question down in the words the desk would use.
The third is the change at the first desk. It should be one question or one check, short enough to ask in every encounter, and explained to the people who will ask it in terms of what it does for the patient or the customer, not for the billing team. Give them a sentence for the patients who ask why. And give them a place to note what they find, because they will begin to find more than the one thing they were asked to look for.
The fourth is the doctors, or whoever in another firm writes the record the payment depends on. Show them patterns, monthly, in a meeting they already attend, never single claims by message. Let them decide how the record changes, and look first for a change to the template rather than to their habits. A practice that turns its clinicians into billing clerks has recovered its revenue by spending its medicine.
The fifth is the write-off. Every claim or deduction written off is recorded with its cause, not as an adjustment, and the total by cause is read once a month. It is the only number in the walk that shows the cost of doing nothing, and most systems hide it by default.
The sixth is the measure. The most useful single figure is the share of claims paid at the first submission, read by cause and by payer, against the practice's own history over a year. It counts denials prevented, which is the gain, rather than denials won on appeal, which is a cost recovered at a cost.
The seventh is the line the practice does not cross, and it needs saying in writing at the start. The walk recovers payment for care actually given and properly recorded, at the rates the practice has agreed. It never adds to the record what did not happen, never chooses a code for its price rather than its accuracy, and never presses a patient for what an insurer owes. A practice should have its billing reviewed by people qualified to do so; nothing in this chapter is billing, coding or legal advice.
Outside medicine the walk is the same. A manufacturer's deductions for short deliveries begin in the warehouse; deductions for late arrival begin in the booking of the carrier; deductions for promotions begin in a conversation between a sales representative and a buyer that nobody recorded. Credit control receives them and works them one at a time. Sorted by cause and walked to the first desk, they turn out, as the practice's denials did, to be a handful of questions nobody asked.
And the decision that most changes the result is that the walk is repeated. Once a month, on whichever pile is largest. The largest pile changes, and the walk goes with it. Chapter 6 places the first walk in the middle of the ninety days, and Chapter 7 gives the sort sheet, the walk sheet, the first-desk question and the monthly board.
It is a quarter past seven on a Wednesday in April, and the billing coordinator is at her desk before the first patients arrive. The week's denials are on her screen. There are seventy-one of them. On the board in the staff room, the coverage pile for the month is a third of the size it was in the autumn, and the front desk has drawn a small line under it, in marker, the way the receptionists asked to.
The first denial of the morning is for a prior authorisation, and she makes a note to walk it on Thursday.
The second is for a woman seen in January for a chest infection, a different woman, with a new job. It is not a denial. It is a payment, from the right insurer, at the first attempt.
Chapter Five
Discounts bought by the calendar, and how a recovered dollar is counted so that everyone believes it.
The Last Week of the Quarter
Haute Lumière · The Price Nobody Charged · Chapter Five
It is nine o'clock on the last Thursday of the financial year, and the regional sales manager of a firm that makes industrial pumps is at her kitchen table with a laptop and a glass of water. On the screen is a deal for a water utility: eleven pumps, a three-year service agreement, and a discount a good deal deeper than the firm's guidance allows. The representative who owns the account rang an hour ago. If the order is signed tomorrow, he reaches the threshold in his commission plan at which every further sale pays at a higher rate, and the region makes its number. If it is signed on Monday, it falls into next year, and neither does.
The utility's buyer knows this. He has had the quotation for three weeks. On Tuesday he told the representative that he could get it approved this week if the price moved, and not otherwise. The sales manager has seen the same conversation at the end of every year she has run the region. Her throat is tight. She approves the discount, writes year-end in the reason field, and closes the laptop.
She is a composite, drawn from the way these evenings are described in the research below and in the house's own practice, and no firm is meant. In the morning the order will be signed, and the firm will report a strong final week. Nobody will report what the final week cost.
Ammanuel Santa Anna's canon, writing about the gap between what organisations say they value and what they reward, puts her evening in two sentences.18 "The real performance system is the one that determines promotions, bonuses, and career trajectories. And that system teaches people, with devastating efficiency, exactly what the organization actually values." The firm's guidance on discounts is on its intranet. The commission plan is in every representative's head. She has just watched the second defeat the first, and she helped.
This chapter makes one claim, and it closes the argument the four chapters before it have made. A great deal of the revenue a firm loses below the invoice is not lost to customers. It is bought by the firm's own calendar, through incentives that make the last week of a period the cheapest time to sell, and it can be recovered only if the people who give the price are paid on what the firm keeps. And any revenue recovered, from this leak or the other four, can be shared only if it is measured in a way that both the firm and the people who found it believe, and that the customer could see without feeling robbed.
The first careful evidence that the calendar moves prices came from an economist looking at whole industries. In 1998 Paul Oyer published a study in the Quarterly Journal of Economics on the effect of fiscal-year ends.19 His starting point was the shape of ordinary pay. "Salesperson and executive compensation contracts typically specify a nonlinear relationship between firm revenues and pay," he wrote, which gives the people paid under them reason to "manipulate prices, influence the timing of customer purchases, and vary effort over their firms' fiscal years." If that were happening, he reasoned, it would show in the data. It did. Manufacturing firms' sales, he found, are higher at the end of the fiscal year and lower at the beginning than in the middle, over and above the ordinary business cycle, and he found further evidence in how prices moved through the fiscal year.
It is evidence at the level of industries and accounts, and it cannot show what any one representative did on any one Thursday. What it shows is that the pattern the sales manager sees in her region is not a local habit. It is visible in the published sales of manufacturing firms across an economy, timed to their own years rather than the calendar's. When thousands of firms pay their people in a way that rewards finishing a period strongly, the whole economy acquires a pulse.
The pulse has a second consequence that the data cannot show directly, and every buyer who has negotiated with a supplier knows it. Customers learn the rhythm. A purchasing manager who has seen a supplier's price soften in the last week of its year three years running does not need a study to tell him when to place his order. He waits. His waiting is not cynical; it is his job, and he is often measured on the savings he finds. So the supplier's incentive plan and the customer's procurement targets lock together, each making the other's behaviour more rational, and the price at which the two firms actually trade drifts down towards the one available in the last week, while both firms' reports describe the deal as normal business.
Oyer's paper showed the pattern. A study published sixteen years later showed its price.
In 2014 Ian Larkin published an analysis, in the Journal of Labor Economics, of the deals done by the sales force of a leading enterprise software vendor.20 The vendor paid its salespeople under a common kind of plan, in which commission rates accelerate as a salesperson's sales for the period rise. Larkin found, in his words, that "salespeople are adept at gaming the timing of deal closure to take advantage of the vendor's accelerating commission scheme." In quarters in which a salesperson had a financial reason to close a deal, salespeople agreed to significantly lower prices. The resulting mispricing, he estimated, cost the vendor six to eight per cent of its revenue. He checked whether the lower prices might simply have been the vendor discriminating, deliberately, between customers, and found that they were not.
It is one vendor, in one industry, with a particular plan. But the finding is unusually direct. It does not say that salespeople are dishonest. It says they are rational, and that a plan which pays more for a deal closed this quarter than for the same deal closed next quarter will buy the deal this quarter, at whatever price is needed. The cost does not appear as a cost anywhere. It appears as revenue that was smaller than it would have been, which is the hardest kind of loss to see.
Read the sales manager's evening through it. The representative was not being greedy. He was being paid. The utility's buyer was not being cynical. He was reading the same calendar. And the firm, which wrote the plan, had in effect offered its customers a discount for buying in the last week of the year, and taught them, year by year, to wait for it. The canon puts the mechanism in a sentence about measurement that applies with full force to pay: "When an organization's reward system is built entirely on quantifiable outputs, people naturally redirect their energy toward what is counted." What was counted was the order. What the firm kept was not.
The representative's Friday
Haute Lumière · The Price Nobody Charged · Chapter Five
It is worth spending a moment with the representative, because the research is easy to read as a story about people who game a system, and he does not experience it that way.
He has a family and a mortgage, and his pay is made of a salary and a commission that, in a good year, is the larger part. For eleven months his plan pays him at one rate. Past a threshold, it pays at a higher one, for everything he sells until the year ends. In the last week of the year he is a few orders short of the threshold, and the difference between reaching it and missing it, on the deals already in his pipeline, is a large share of what his family will live on. The utility's order is the one that would take him over.
He does not think of the discount he asked for as costing the firm anything. He thinks of it as the price of getting a deal done that the firm wants done. He has been told, at every sales conference, that the firm wants volume, and the plan he is paid under agrees. Nobody has ever shown him what the firm keeps from a deal, and his plan does not ask. If the firm wanted him to hold the price, he would say, if anyone asked him, it would pay him to hold the price.
He is right, and that is the whole difficulty. The research does not describe salespeople behaving badly. It describes people paid to do one thing doing it well. Larkin's salespeople were adept, which is a compliment. The sales manager who approved the discount at nine on Thursday was doing the same thing at the next level up: her own plan rewarded the region's number, and the number needed the order. A firm that finds its pocket price eroded at every quarter's end has not found a discipline problem in its sales force. It has found its own pay plans, working as designed.
That is also why exhortation does not change it. A sales director who tells the team at the start of the year that discounting must stop, and leaves the plans as they are, will hear the same phone calls in the last week. The behaviour is not a habit to be broken. It is an answer to a question the firm keeps asking.
There is an older question beneath the calendar, and it is the question of who is allowed to set the price at all.
In 1979 the marketing researchers P. Ronald Stephenson, William Cron and Gary Frazier published a study, in the Journal of Marketing, of a question that the textbooks of the day answered one way: should salespeople be given authority to set prices? The theory said yes, because the salesperson knows the customer best. Their data, from a sample of a hundred and eight firms, said otherwise. Among those firms, the ones "giving salespeople the highest degree of pricing authority generated the lowest sales and profit performance."21
It is an old study and a correlational one. It cannot show that the authority caused the poor performance; firms that were struggling might have handed their salespeople more room to discount, and the relationship might differ in other industries and decades. It is cited here not as a law but as a caution with a long history. The person who knows the customer best is often also the person most exposed to the customer's pressure, and most rewarded for giving way to it. Giving that person the whole of the pen, and paying them on volume, puts both forces on the same side of the table.
Nothing in this argues for taking pricing away from the people who sell. They do know the customer, and a firm that prices from a distance prices badly. It argues for something simpler: that the person who grants a discount should see, at the moment of granting it, what the firm will actually keep, and should be paid in a way that cares about the answer. The discount the sales manager approved at nine on Thursday went into a field marked reason. Nobody at the firm will read it, and the commission plan does not know it exists. The house's guide to selling, in a chapter on the conversation about price, lists among the skills it teaches: "When to walk away: recognizing that price is a genuine barrier vs. a negotiation tactic". The utility's buyer was using a tactic. The firm's own plan made it impossible for the representative to recognise one.
Santa Anna's Ecstatic Economics, writing about the gift economies it reads as models for a living firm, describes wealth in terms that fit a quarter-end surprisingly well: "In the gift economy, wealth is a flow — a current of energy that must keep moving to remain alive." The quarter-end is a dam. Orders that would have arrived steadily through the year are held back until the last week, then released all at once, at a lower price, by customers who have learned that holding them back pays. The firm's production plans see a wave. Its service teams see a wave. Its margins see a trough. None of this was anyone's intention. All of it was designed.
The house's guide for customer-service teams makes a related point in two short sentences that a sales director could keep above her desk: "The metrics flow from the states. Not the other way around." It was written about the state of mind in which people serve customers, and it holds at a quarter-end. A representative in the last week of the year is in the state of someone chasing a threshold, and the price follows the state.
The canon is careful to say that the remedy is not to stop measuring. The same chapter that describes the real performance system lists, among the harms of a reward built on short-term counts, the habit of "prioritizing short-term metrics over long-term relationships". The quarter-end discount is that habit in its most concentrated form. It trades the price of a three-year relationship for a number in one week.
It is worth naming who bears the cost, because it is not only the firm. The representatives carry the last week of every quarter in their bodies, and the best of them leave for firms where the year does not end in a scramble. The customers who do not know the calendar, often the smaller and more loyal ones, pay more than the ones who do. And the band that a commercial analyst draws for a single product (Chapter 1) widens a little every quarter, as the customers who wait move to the bottom and the customers who do not stay at the top.
What changes it, in practice, is smaller than a new commission plan, though a new plan may follow. At the pump firm the sales director makes three changes for the following year. Every discount beyond guidance is written on the deal record with its reason, and the record shows the pocket price the firm will keep, worked out from the same ledgers the waterfall uses, before the discount is approved. Discounts granted in the last fortnight of a period are reviewed by someone whose own pay does not depend on that period. And the representatives' commission is counted on what the firm keeps from each deal as well as on what it ships, so that a deal signed on Monday at the right price is worth as much to the representative as the same deal signed on Friday at the wrong one. How any of that is designed for a particular firm, the rates and the thresholds and the timing, is a decision for that firm and its people, and it is not described here.
The first year is uneven. Two large customers wait for the old year-end and are surprised when it does not come. One of them signs in the first week of the new year at the guided price. The other takes its order to a competitor, and the sales director lets it go, which is the hardest decision of the year and the one the representatives notice most.
A number the sales team believes
Haute Lumière · The Price Nobody Charged · Chapter Five
Every leak in this book ends in the same place. The commercial analyst finds a waterfall, the engineer a ledger of favours, the commercial manager a drawer of clauses, the billing coordinator a pile of denials that began at the front desk (Chapter 4), the sales director a calendar. Each finds revenue the organisation has already earned and has not collected. And in each case the finding becomes a gain only if it can be counted in a way that the people who found it, the people who must pay for it and the customers who are affected by it would all accept.
Where the Flow Stops reads a measurement agreed before anyone knows the answer, and the principles are the same here. A baseline both sides lock at the start, drawn from records the organisation already keeps, over a period long enough to show its ordinary variation: for revenue, the pocket price per unit and the share of work billed, read from the waterfall's ledgers and the timesheets over a year or two rather than a single bad quarter. Adjustments named in advance for the things that move revenue for reasons nobody in the firm controls: the mix of products sold, the volume, a published index of input costs, the market's own movement in price. A calculation done or checked by someone with no stake in the answer. And the result shown to the people whose work produced it.
Revenue adds one requirement that the measurement of costs does not, and it is the most important line in this chapter. A revenue gain must carry a second measure beside it that would reveal if the gain were merely taken from customers. Anderson and Simester's customers, whose purchases fell when they learned they had paid more than others, are the warning.22 A firm that recovers its pocket price by squeezing the customers least able to argue will see it first in their volumes and their departures, a quarter or two later. So the recovered revenue is read beside customer retention, the volumes of the firm's smallest and most loyal accounts, and the number of disputes. If those hold steady while the pocket price rises, the firm has recovered what it was owed. If they fall, it has moved its costs onto other people and called it a gain, and the measurement should say so.
Consider how this would work at the pump firm, told as the sales director and the finance director agree it in the first week of the year, before anything changes. They take two years of the firm's deals for its main pump lines, from the deal records and the waterfall's ledgers, and work out the pocket price per pump, month by month, so that the ordinary swing of the year, including every past quarter-end, is visible. They name the adjustments in advance: a published index of steel and copper prices, because the firm's list prices move with it; the mix between large and small pumps, because the large ones carry different discounts; and volume, because a slow year changes what customers can demand. They name the balancing measures: the number of active customers in each region, the volumes of the smaller accounts, and the deals lost at the quotation stage, recorded with the reason the customer gave. The management accountant, who reports to finance and is paid on neither sales nor margin, will run the figures each month, and the sales team will see them in the same form the finance director does.
At the end of the year the pocket price has risen against its two-year range, after the adjustments. Active customers have held. One large account has been lost, and is shown, with the revenue it took with it, beside the gain, so that nobody can claim the gain without seeing the loss. The figure that results is smaller than the sales director's estimate in the spring. It is also a figure the representatives believe, because they watched it built from records they recognise, every month, and saw their own deals in it.
This is not decoration. It is the difference between recovering revenue and extracting it. Every practice in this book recovers money from places where the organisation was simply failing to collect what it had agreed or earned: a rebate paid against volume that was not reached, a site visit the client was never told about, an indexation clause in a signed contract (Chapter 3), a claim for care already given, a discount bought by the calendar. None of them requires a customer to pay more than it agreed. Several of them, the installer's price, the released years of indexation, the named gift, return something to the customer. A gain measured with its balancing measure beside it can show that, and a gain shared on that basis can be defended in public.
Chapter 7 carries the measurement conversation for revenue as a run sheet, with the questions in order, the adjustments to name and the balancing measures to set beside each kind of leak, and the section below takes its finer points further. It does not describe how a measured gain is divided, which belongs to each organisation and the people it covers.
The house's interest in this is stated once, plainly. Luminous Prosperity works with organisations to find the revenue and the gains they already hold, and to share in what is found. Sharing is only honest when the measurement is, which is why the measure is agreed first, and read last.
The measurement of recovered revenue follows the same order as the measurement of any gain, and Where the Flow Stops sets that order out in full. Revenue has finer points of its own, and they are the ones that decide whether the number is believed.
The first is the unit. Revenue rises and falls with volume, and a firm that measures recovery in total revenue will credit itself with a good year and blame itself for a bad one. Measure the thing each practice changes. For the waterfall and the incentives, the pocket price per unit of a product line. For the scope ledger, the share of hours worked on a project that were either billed or deliberately given. For the register, the revenue billed under each clause against what the clause provided. For the denial walk, the share of claims or invoices paid in full at the first attempt.
The second is the history. Copy and date at least a year of the records that bear on each unit before anything changes, and two years where the business is seasonal or the quarter-end distorts it. The history shows the ordinary range, and nothing is counted as recovered until it falls outside that range or moves steadily in one direction for several months.
The third is the adjustments, and they are named before anyone knows the result. For price: a published index of the firm's main input costs, if its prices move with them; the mix of products; volume; and the market's own price movement, where a source for it exists. For billed hours: the mix of fixed-fee and time-based work. For denials: changes in the payers a practice works with and in their rules. Each adjustment names its record and how it will be calculated. An adjustment proposed after the results are known is not an adjustment. It is an argument.
The fourth is what is not counted. Revenue released on purpose, the missed years of indexation, the named gift, is counted as zero, never as a gain foregone. Revenue that comes from charging a customer more than it agreed is not counted at all. Changes to incentive plans are not gains in themselves; they are counted only through the pocket price they move. And recovered amounts are counted from the date the change was agreed, not backwards.
The fifth is the balancing measure, which revenue needs more than any other gain. Beside every revenue measure sits a measure that would rise or fall if the gain were taken from customers rather than recovered. The number of active customers. The volume bought by the smallest and most loyal accounts. Quotations lost, with the reason given. Disputes, complaints and, in a practice, bills sent to patients. If these hold while the revenue measure improves, the firm has collected what it was owed. If they deteriorate, the result is reported beside the gain, in the same table, so that nobody can read one without the other.
The sixth is who calculates. Someone whose pay depends on neither the sales figure nor the margin, with a second person able to check the work. The result is shown monthly to the people whose practices produced it, the account managers, the engineers, the front desk, in the form finance sees.
And the seventh is what this conversation does not settle, which is how a measured gain is then shared: the proportions, the timing, the thresholds and the terms. Those depend on the organisation, its people and its history, and for an organisation working with Luminous Prosperity they are the substance of the engagement, not of a book. What the conversation can do is make that later decision honest, by producing a gain that the firm, the people who found it and the customers it affects could all look at. Chapter 6 places the conversation at the end of the ninety days, with its history fixed at the beginning, and Chapter 7 gives the questions in order and a one-page agreement.
It is nine o'clock on the last Thursday of the financial year, and the regional sales manager is at her kitchen table with a glass of water. The laptop is closed. The region made its number in the second week of the month, without the last week, and two of the deals that would once have waited for Friday were signed in August at the guided price. The water utility signed its renewal in July. The buyer asked for a discount, and was shown, on one page, what the firm would keep and why the price was what it was. He agreed to the price, and asked for a longer service agreement instead.
The representative who rang her last year at eight is on holiday this week. The reason field on the region's deal records holds sentences now, not words.
She drinks the water, and goes to bed early.
Chapter Six
Five practices set in order, from fixing the history before anything changes to measuring what was recovered.
The First Ninety Days
Haute Lumière · The Price Nobody Charged · Chapter Six
It is a quarter to eight on a Monday, and the finance director of a firm of some three hundred people has the first clear morning she has had since the year-end. The firm makes and services equipment for commercial kitchens. It sells through account managers to restaurant groups and hotels, and directly to a long tail of small independent restaurants; it runs a service division that bills engineers' time; and it holds a few dozen supply and maintenance agreements with larger customers, some of them many years old. She has read the first five chapters of a book like this one, or lived them, and she suspects that every one of the five leaks is somewhere in her ledgers. She has a notebook, a coffee, and a list of the firm's twenty largest customers. She does not yet have a plan, and she can feel the pull, in her shoulders, to write one by lunchtime and announce it on Friday.
She is a composite, drawn from the way these first weeks are described in the research the earlier chapters read and in the house's own practice, and no firm is meant. This chapter is written to her, and to anyone in her position: the person who runs something, or runs its money, and wants to begin. It describes the first ninety days in order, with who is in the room, what each step needs, roughly how long it takes, the decisions that change the result, the mistakes made at each step and how to see them early, how to read what comes back, and what to do on the first Monday. The worksheets and run sheets it refers to are gathered in Chapter 7.
Two things should be said before the plan. The first is that nothing here requires a programme, a consultant's software or a name. Ammanuel Santa Anna's canon, writing about the relational side of how value moves through an organisation, gives the warning in six words: "Relational value is not a program."23 A finance director who launches the practices in this book as a revenue-assurance initiative, with a title and a slide, will find that account managers respond to the initiative, defensively, and not to the practices. The second is that the order matters more than any single step. The chapters were written in the order a reader meets the leaks. The ninety days run in the order the practices depend on each other.
The five practices of this book are the waterfall and the band (Chapter 1), the scope ledger (Chapter 2), the clause register and the call (Chapter 3), the denial walk (Chapter 4), and the discount record with the measurement conversation (Chapter 5). In the first ninety days they run in this sequence: fix the history, see the price, read the contracts, walk the work, speak to customers, and only then change incentives and agree the measure.
The reasoning has three parts. First, every practice in this book eventually changes something a customer sees, and the measure of what was recovered is only believable if it is read against a history fixed before anything changed. So the first move, on the first days, is to copy the records and date them. Chapter 5 explains why, and Where the Flow Stops explains it at greater length: a baseline cannot be re-created once the work has moved it.
Second, everything is seen inside the firm before anything is said outside it. The waterfall, the register and the walk are all internal readings. They can be done without a single customer knowing, and they should be, because the conversations of Chapter 3 go well only when the firm knows exactly what it is asking, why, and what it is giving back. A firm that rings its customers in week two, before it has drawn its own waterfall, will ask for the wrong things and concede the wrong ones.
Third, incentives come last. It is tempting to start with them, because the quarter-end discount of Chapter 5 is so visible. But a pay plan redesigned before the firm has seen its waterfall is redesigned on guesswork, and the account managers will experience it, correctly, as a judgement made about them without evidence. Once the waterfall and the register are on the table, the case for changing what the plan counts is in front of everyone, in the firm's own numbers.
The High Frequency Workbook says of a person who has avoided her own accounts for years something that every firm beginning this work should hear: "Financial avoidance is not laziness — it is a protection strategy. Your system is avoiding information that it anticipates will be painful. This chapter is about making it safe enough to look." So are the first thirty days.
Three roles carry the first ninety days, and they should be named before anything else is done.
The sponsor is the most senior person whose authority the practices will need: usually the chief executive or the finance director, and in a large firm the head of a division. The sponsor's job is not to run the work. It is to protect the people who will find uncomfortable things, above all the account managers and the billing staff, and to make it plain, early and often, that the exercise is about the system and not about anyone's past decisions. Expect it to take three to four hours a week.
The working group is small: five or six people. Someone from finance who knows where the ledgers are. A commercial lead who can read a contract. One experienced account manager, chosen because colleagues trust her rather than because she is senior. Someone from billing or credit control, who sees the denials, the deductions and the late payments every day. Someone from operations or service, who knows what is delivered that never reaches an invoice. The group meets for an hour a week.
The convenor keeps the record: the copied histories and their dates, the waterfalls, the register, the walk sheets, the decisions taken and why. Often this is the finance member. The convenor needs about three hours a week in the first month and two thereafter, and the quality that matters most is that the account managers trust the convenor not to turn what they say into a case against them.
One decision changes the result more than any other at this stage. The sponsor must decide whether she will sit with an account manager, at least once, while a waterfall is drawn, and hear from him what each concession was for. A sponsor who reads waterfalls only in summary will see a list of leaks. A sponsor who sits through one will see a list of reasons, which is what the firm actually has to change.
The first ten days change nothing visible. They gather five inputs, and fix one of them in place.
The first is the history. Before day eleven, finance and the convenor copy, date and file, unaltered, at least a year of every record the later practices will read, two years where the business is seasonal: invoices and the price master; rebate files and rebate payments; credit notes and returns; freight and delivery charges; payment days actually taken, by customer; timesheets and project budgets for the service division; deal records with their discounts and the dates they closed; and, if the firm has them, denials, short payments and customer deductions with their reasons. Nothing is decided about any of it. It is the history the measurement conversation will read on day seventy, fixed before any practice could move it.
The second is a list of ledgers and their owners. For one product line and one customer, the convenor writes down every place a piece of the price is decided or recorded, and who keeps it. The list usually runs to seven or eight entries and five or six departments. It is the first finding, and it should be shown to the working group without comment.
The third is a choice of customers: three for the first waterfalls, chosen as Chapter 1 describes, one large, one ordinary and one small and loyal, and one product line for the band.
The fourth is a list of the forty largest customers' agreements and where each is kept, without yet reading them. The commercial lead estimates how long the reading will take, and the estimate is usually doubled.
The fifth is one week of the denial or deduction queue, or, in a firm without one, of credit control's write-offs, simply counted by the reasons already recorded. The point is to know how large the pile is and roughly what it holds before anyone walks it.
At the end of the ten days the working group meets for the first time and writes, on one page, its first guess at where the revenue leaks: which bars of the waterfall it expects to be widest, which contracts it suspects, where it thinks work goes unbilled, what it expects the queue to show. It is a guess, and labelled as one. The page is kept and read again on day ninety.
What goes wrong at this stage. The commonest mistake is to announce. A sponsor who tells the sales team that finance is about to review discounting will find that every concession granted in the next month comes with a careful reason attached, and that the account managers stop speaking to the working group. The early sign is a question in a corridor: is this about our commission? If people are asking, it has been announced, and the sponsor should stop describing it. The second mistake is to begin with software. Revenue-assurance tools exist and can be useful later; bought in week one, they replace the reading the firm needs to do with a dashboard nobody yet knows how to read. The third mistake is to skip the history because the records are untidy. Untidy history, copied and dated, is worth far more than tidy history reconstructed after the work has begun.
Days eleven to thirty: seeing the price
Haute Lumière · The Price Nobody Charged · Chapter Six
In the third week the finance member and the account manager on the working group draw the first three waterfalls by hand, each in a two-hour session, with the account manager responsible for that customer in the room. Chapter 7 gives the sheet. Each bar is drawn from its own ledger, with its source written beside it, and then each bar is asked the three questions of Chapter 1: who approved it, and when; what was the reason at the time; does the reason still hold?
In the fourth week the convenor works out the pocket price for every customer who bought the chosen product line in the year, and draws the band. The working group spends an hour with it, asking what explains the order: volume, cost to serve, the length of the relationship, the account manager. What no reason the firm would defend can explain is written down as the finding.
The decisions that change the result are about tone and about the small customers. Every bar is given one of three provisional answers, keep, renegotiate or stop, but nothing is changed and nobody outside the room is told. And the customers at the top of the band, the ones who may be paying more than their business costs the firm, are listed as carefully as those at the bottom. The working group should expect the list of prices to come down to be shorter than the list to go up, and should treat it as just as important.
What goes wrong. The waterfall is drawn without the account manager, and becomes an audit of his judgement. The band is read only from the bottom. A finding is shared with the sales director before the account manager concerned has seen it. Freight and returns are spread as averages, and the differences the band exists to show disappear. The early signs are visible in the room. If the account manager defends every bar, he has been made to feel accused, and the sponsor should say, in a sentence, what the exercise is for. If nothing on the band surprises anyone, it has probably been drawn from averages.
How to read what comes back. By day thirty the convenor's record should hold three waterfalls with an answer beside every bar, a band with its unexplained spread marked, and a short list of the concessions whose reasons have expired. The bars that recur across all three customers are the firm's habits rather than any account's history. They are where to look first.
Overlapping with the waterfalls, the commercial lead and an administrator read the forty largest agreements for the clauses that move money, and build the register Chapter 3 describes. The reading takes, in most firms, a fortnight of one person's time spread over three or four weeks.
For each clause the register records its next date, the person who will act on it, and what was done last time it fell due. The commercial lead notes every place where two documents disagree, the purchase order and the acknowledgement, the agreement and a later letter, because the disagreement is itself unfinished business. Each unused clause is given one of the three decisions of Chapter 3, apply, revise or release, with its reason, and, as with the waterfall, nothing is said to any customer yet.
The decision that changes the result is the default for the past. The working group recommends, and the sponsor decides, a rule for missed years: in most cases they are released, openly, as the firm's own oversight, and the clause is applied from its next date. Exceptions are taken one at a time, with the firm's advisers, and nothing in this chapter is legal advice. A firm that decides this rule before the first call will make every call more easily, because the person making it knows what she can offer.
What goes wrong. The register is built by the legal team alone and becomes a list of rights rather than a calendar of decisions. Every unused clause is marked apply, including the ones that no longer describe the customer's business. The register is built and never read again. The early sign is the register's own dates: if the first date passes after day forty-five without a name beside it and a note of what was done, the register has become a file.
How to read what comes back. By day forty-five the register should show, for each of the forty agreements, the clauses that move money, their next dates, and a decision against each unused one. The total of what applying the apply clauses would bring from their next dates, with the release decisions shown beside it at nothing, is the first estimate of this leak. It is an estimate until the calls are made.
In the second month two practices begin, side by side, in the parts of the firm that do the work.
In the service division, the scope ledger of Chapter 2 starts on two live jobs: one maintenance contract and one installation. The engineers write a line on the day a customer asks for something the job does not include, and the service manager answers every line on Mondays: change, given or declined. Before the ledger starts, the service manager writes down, with the engineers, the three or four requests that most often fall outside a job of each kind. In a firm without a service division or professional work, the same ledger can be kept by whoever handles special requests: same-day deliveries, extra visits, rush orders, samples.
In credit control, the denial walk of Chapter 4 is run on the last full month of short payments and deductions. The pile is sorted by cause, three ordinary items from the largest pile are walked backwards to the desk where the cause began, and one question or check is agreed at that desk. In a medical practice it is the denial queue; in a manufacturer it is deductions for short deliveries, late arrivals and disputed promotions; in a professional firm it is the invoices the client disputes.
The decisions that change the result are, for the ledger, the first given, and for the walk, the first change at the first desk. The first gift should appear on an invoice as a line at no charge, within a month, so that the engineers see the ledger produce a thank-you as well as a charge. The first change at the first desk should be one question, explained to the people who will ask it in terms of what it does for the customer.
What goes wrong. The ledger asks for too much and is filled in on Fridays from memory. Every line is answered given. Customers learn about changes only from their invoices. The walk is run on the largest single item rather than the largest pile. The first desk is told about the walk as a complaint. The early signs: a ledger with fewer lines in week four than in week one has stopped being kept; a first desk that asks why it is being blamed has been told about the walk in the wrong way.
How to read what comes back. By day sixty the ledger should show a handful of lines a week on each job, with their answers, and a first view of how much work falls outside the scope of a typical job. The walk should show the share of the month's pile that began at one desk, and the question now asked there. Neither will yet show a gain. Both show where it is.
With the waterfalls drawn and the register built, the firm begins the conversations, in the order of the register's dates and the customers' own negotiation calendars, never all at once.
Each conversation is made by the person who holds the relationship, with the commercial lead or the finance member beside them for the first few. It follows the order Chapter 3 gives: name what the firm found and when; say what the firm did not do and that the missed period is its own to absorb; say what will happen from now and why; ask about the customer's side; offer what the waterfall showed in the customer's favour; confirm it in writing the same week; and only then change the invoice. For the small customers at the top of the band, the conversation is shorter and easier: the firm has found that it was charging more than the business warranted, and it is bringing the price down.
The decision that changes the result is to let a few conversations end without agreement. Some customers will refuse a change, and some will threaten to leave. The sponsor decides in advance which concessions the firm will hold to even at the risk of an account, and which it will trade, so that the person on the phone is not deciding alone, under pressure, what the firm is worth.
What goes wrong. The calls are made by finance rather than by the person who holds the relationship. The first a customer hears is a letter. Several changes are put to one customer at once, by surprise. The account manager trades away the change in the next conversation, because nothing about his pay has changed yet. The early sign of the last of these is visible in the waterfall itself, redrawn a month later: bars that were closed are open again under another name.
How to read what comes back. By day seventy-five the convenor's record should show, for each conversation, what was proposed, what was agreed, what was released and what was refused. The agreed changes, from their effective dates, are the first recovered revenue. The released items are the firm's gifts, now visible. The refused ones are information about which concessions the market actually requires, and they belong in the next year's price list.
Days sixty-one to ninety: the incentive and the measure
Haute Lumière · The Price Nobody Charged · Chapter Six
In the last month the sponsor and the sales director take the step Chapter 5 describes, and the working group holds the measurement conversation.
The step on incentives has three parts, and only the first two need to happen in the ninety days. Every discount beyond guidance is recorded on the deal with its reason, and the record shows the pocket price the firm will keep, worked out from the same ledgers the waterfall used, before the discount is approved. Discounts granted in the last fortnight of a period are reviewed by someone whose own pay does not depend on that period. And the sales director begins the design of a plan that counts what the firm keeps as well as what it ships, a design that belongs to the firm and its people and is not described here.
The measurement conversation follows the order Chapter 5 and Chapter 7 give: the unit for each practice; the history fixed on day ten; the adjustments, named now, before any result is known; what is not counted; the balancing measures; who calculates; how the result is shown. Finance leads it, the commercial lead and an account manager are in the room, and so is someone from credit control. The output is one page, signed by the people in the room. It usually takes two meetings of ninety minutes.
By day ninety the first monthly figures should be on a page the sales team and the service engineers can see, plotted against the ordinary range of the fixed history, with the balancing measures in the same table.
What this step does not settle is how a measured gain is then shared with the people who found it: the proportions, the timing, the thresholds and the terms. Those depend on the organisation, its people and its history. They are decided with the people who will be covered, and for an organisation working with Luminous Prosperity they are the substance of the engagement, not of a book. What the first ninety days can do is make that later decision honest, by producing a gain the firm, its people and its customers could all look at.
What goes wrong. The baseline is drawn from the worst recent period, because it makes the recovery look large. An adjustment is added after the results are in. Revenue released on purpose is counted as a gain foregone. The balancing measures are left off the page that the sponsor reads. The early sign is the same in every case: someone in the room, often the account manager, goes quiet. The sponsor's job at this step is to notice who has stopped speaking.
The working group meets on day ninety with six things on the table: the first-guess page from day ten, the waterfalls and the band, the register, the scope ledgers, the walk sheets, and the signed measurement page with its first figures.
It asks a short list of questions, in order. Where was the first guess right, and where wrong? Which bars of the waterfall were closed, and which reopened? Which clauses were applied, revised and released, and what did the customers say? What share of the work outside scope was billed, what share given on purpose, and what share is still invisible? What share of the walked pile began at one desk, and has it moved? What does the revenue measure show, and what do the balancing measures show beside it? And then the question that matters most for the next ninety days: which customers paid less as a result of this work, and do they know?
What not to conclude is as important. Ninety days are too short to say the leaks are closed; they are long enough to say whether the practices are being kept. A single month's rise in pocket price is not yet a gain, for the reasons Chapter 5 gave. And a quiet register or an empty ledger is a finding in itself. It usually means a practice has become a file, and the first place to look is whether anyone with authority has read it this month.
The canon's chapter on governance describes the regular check on commitments that keeps this alive, and the tone it asks for: "Not a performance — a report." The day-ninety review is that report, for the firm's revenue. Nobody needs to be praised for it and nobody blamed. It says where the money went, and what the firm decided about it.
If the whole of this chapter is too much to begin with, three things can be done on the first Monday, and none of them needs anyone's permission.
Choose the customer you think of as your best, and write down, for the product it buys most, every bar between the list price and what stays in the firm's pocket. If you do not know where one of them is recorded, write that down too.
Open the agreement with your second-largest customer, and find the next date on which a clause that moves money falls due. Put it in a diary, with a name.
And ask, once, the question the High Frequency Workbook puts to anyone avoiding their accounts, which comes in the first step of its exercise on knowing your numbers: "Look at the balances. Write them down."
The canon on keystone work begins its account of what to do with what is hidden from view with one plain sentence: "The first and most foundational principle is visibility." Every practice in this book is a way of making one kind of revenue visible to the people who decide what to do with it. The leaks were never secret. They were simply not looked at.
It is a quarter to eight on a Monday, and the finance director's diary is full again. In a folder on her shelf are three waterfalls drawn in pencil, each with an answer beside every bar. The register on the shared drive shows four dates this month, each with a name. In the service division, two scope ledgers lie open on the engineers' bench, and one of last week's invoices carries a line at no charge that the customer mentioned in an email. On the wall in credit control, the deductions for short deliveries are counted by cause, and the warehouse's pile is smaller than it was in the spring. On the sales director's screen, every discount beyond guidance has a sentence beside it, and a pocket price.
None of these has a name. Nobody has announced anything. The finance director opens the register, notes the first date of the week, and closes the laptop.
Chapter Seven
Sixteen working pages, the waterfall, the ledger, the register, the walk, the discount record and the agreement, ready to carry into a room.
The Field Manual
Haute Lumière · The Price Nobody Charged · Chapter Seven
The chapters of this book describe five practices as people doing them.24 This chapter holds the working pages: the sheets, the run sheets, the question sets and the one-page templates that the earlier chapters point to and do not print. Each page is meant to be photocopied, filled in by hand, and kept in a folder that the convenor of Chapter 6 looks after. Nothing here needs software, though most of the figures come from systems the firm already runs.
The pages follow the order of the first ninety days, not the order of the chapters. Each opens with a line saying where it comes from and when it is used. Where a page gives a script, the words are a starting point; say them in your own voice. Where it gives a time, the time is a planning guide from practice, not a finding.
Three rules govern the whole manual. Write down what people said, not what you wish they had said. Nothing changes for a customer before the customer has been told. And count what was given back as carefully as what was recovered.
What these pages do not contain is any price, any rate, or the design of any arrangement for sharing a gain: proportions, thresholds, timing or terms. Those belong to each organisation and the people it covers, and they are decided separately. Nothing here is legal, financial, billing or accounting advice.
From Chapter 5 and Chapter 6. Completed before day eleven.
Copy each record, date the copy, and file it unaltered. Do not clean it. Note what is missing.
| Record | Period copied (from–to) | Where it lives | Who keeps it | Date copied | Gaps noted |
|---|---|---|---|---|---|
| Invoices and the price master | |||||
| Rebate files and payments | |||||
| Credit notes and returns | |||||
| Freight and delivery charges | |||||
| Payment days actually taken, by customer | |||||
| Timesheets and project budgets | |||||
| Deal records with discounts and close dates | |||||
| Denials, short payments, deductions, with reasons | |||||
| Write-offs, with reasons |
Reading it. At least a year; two where the business is seasonal or periods end in a rush. A gap is not a reason to stop. It is a reason to note the gap and copy what exists.
From Chapter 1, after Marn and Rosiello (1992). Used in days eleven to thirty, for three customers.
Customer: Product line: Year: Account manager present:
Work per unit of the product. Write the source of every bar.
| Step | Amount per unit | Source (ledger and who keeps it) |
|---|---|---|
| List price | ||
| less discounts on the invoice | ||
| = Invoice price | ||
| less early-payment discount (as taken) | ||
| less annual or volume rebate | ||
| less promotional or marketing contributions | ||
| less freight and delivery paid by the firm | ||
| less cost of returns and credits | ||
| less anything else agreed off the invoice | ||
| = Pocket price |
Allocating. Spread freight by the orders that incurred it and returns by the lines returned. Never by an average across customers.
Reading it. The distance from invoice price to pocket price is the part of the price nobody in the firm owns. The pocket price is not the profit on the account: cost to serve differs. Note what the firm knows about cost to serve for this customer, and where it knows little.
From Chapter 1. Asked with the account manager, bar by bar, after page 2 is drawn.
| Bar | Who approved it, and when? | What was the reason at the time? | Does the reason still hold? | Keep / renegotiate / stop |
|---|---|---|---|---|
In the room. The account manager supplies what the ledgers cannot: what was asked, what the alternative seemed to be. Every bar was a reasonable answer to a real question. Ask about the reason, never about the mistake.
Nothing is changed yet. The last column is provisional until page 5.
From Chapter 1. Used in the fourth week, for one product line.
| Rank | Customer | Pocket price | Volume | Cost to serve (known / estimated / unknown) | Terms last set | Account manager |
|---|---|---|---|---|---|---|
Reading it. Ask what explains the order. Volume and cost to serve should explain some of it. Whatever no reason the firm would defend can explain is the finding. Mark the customers at the top who may be paying more than their business costs the firm, and list them as carefully as those at the bottom.
5 · The customer conversation
Haute Lumière · The Price Nobody Charged · Chapter Seven
From Chapters 1 and 3. Used from day forty-six, in the order of the register's dates and the customers' own negotiation calendars.
Who calls. The person who holds the relationship. For the first few, the commercial lead or finance member is beside them.
Before the call (fifteen minutes). One sheet showing what the firm found, what it proposes, what it is releasing and what it is giving back. Know in advance which changes the firm will hold to if the customer refuses, and which it will trade. The sponsor decides this, not the caller.
The order. Do not skip ahead.
Mistakes to watch for. A letter before a call. Several changes at once, by surprise. Back-billing by default. Finance calling instead of the relationship holder. Trading away the change in the next conversation.
From Chapter 2. The list is written before the ledger starts; the ledger is kept on each live job or project from day thirty-one.
The scope list. Read the appointment or statement of work with the team, and write on the first page, in plain words:
The canon's guidance on mapping work puts the reason for this page in a phrase that applies to any engagement: "Clarity here prevents the common failure of scope creep".
The ledger. One line per request, written on the day it is made, by the person who receives it. A minute, no more.
| Date | Who asked | What for | Rough time | Answer (change / given / declined) | Answered by, on | Client told on |
|---|---|---|---|---|---|---|
From Chapter 2. Every Monday, fifteen minutes, the project lead with whoever can commit the firm.
For every new line, one answer.
The sentence for a change. You asked us on (date) to (request). That falls outside what we agreed for this stage. It will take about (time). Would you like us to go ahead? Where the work had to be done at once, say so the same day and confirm in writing.
Signs it is failing. Lines written on Fridays from memory. Every line answered given. Clients told only by invoice. Fewer lines each week.
When writing the next fee. Read the ledgers of the last few comparable jobs first. Count the lines and what they took.
The utilisation report. Show hours absorbed on jobs over budget as their own column, so that the person who absorbed most is seen as the person who did most.
From Chapter 3, after Macaulay (1963). Built in days twenty-one to forty-five for the forty largest agreements; read monthly thereafter.
What to gather for each customer. The agreement, its schedules and amendments; the standard terms each side claims, on purchase orders and acknowledgements; any letter that changed them. Note every place two documents disagree.
What to read for. Only clauses that move money:
The register page.
| Customer | Clause | What it provides | Next date | Who acts | Done last time | Decision (apply / revise / release) and reason |
|---|---|---|---|---|---|---|
The calendar. Every next date goes into the named person's diary, with a month's notice.
From Chapter 3. One decision for every unused clause, written down.
| Decision | When it fits | What happens |
|---|---|---|
| Apply | The clause still describes the business, and the customer knew it would be used | Applied from its next date, after the call (page 5) |
| Revise | The clause no longer fits: the index has drifted, the volume was set for another business, the terms are out of line with the market | A change is proposed in the call |
| Release | The relationship, the history or the firm's own oversight makes it wrong to use, for a stated period | Written down with its reason and its period; the customer is told |
The past. The working rule, decided by the sponsor before the first call: missed periods are released, openly, as the firm's own oversight, and the clause applies from its next date. Exceptions are taken one at a time with the firm's advisers.
A release in writing. Under clause (number) of our agreement, (what it provides) was due on (dates) and we did not apply it. We will not apply it for that period. From (date) it will apply as the agreement provides.
Monthly. Half an hour, the commercial lead and the finance director: what falls due in the next sixty days, and who is calling whom.
From Chapter 4. The last full month of denials, short payments or deductions, from day thirty-one; repeat monthly on the largest pile.
The sort. By what would have prevented it, not by the payer's reason code.
| Cause | Count | Value | Notes |
|---|---|---|---|
| Who is covering or paying (eligibility, the wrong account) | |||
| Permission needed in advance and not obtained | |||
| What the record said or failed to say | |||
| Deadlines missed | |||
| Delivery, quantity or timing disputed | |||
| Everything else |
The walk. Three ordinary items from the largest pile. Walk each backwards from the denial to the first contact. At each desk:
| Desk | What did it know? | What did it not know? | What single question would have closed the gap? |
|---|---|---|---|
11 · The first-desk question and the board
Haute Lumière · The Price Nobody Charged · Chapter Seven
From Chapter 4. Agreed after the walk.
The question, in the desk's own words:
The sentence for anyone who asks why: (in a practice, for example: so that the right insurer pays and nobody receives a bill by mistake)
A place to note what the desk finds. A sheet beside the desk, collected weekly.
The board. Each week, the count by cause, where every desk can see it. Beside it, the share of claims or invoices paid in full at the first attempt.
Write-offs. Each one recorded with its cause, never as an adjustment. Read the total by cause monthly.
The line not crossed. The walk recovers payment for what was actually done and properly recorded, at the rates agreed. It never adds to a record what did not happen, never chooses a code for its price rather than its accuracy, and never presses a customer or patient for what someone else owes.
From Chapter 5, after Oyer (1998), Larkin (2014) and Stephenson, Cron and Frazier (1979). From day sixty-one.
For every discount beyond the firm's guidance, before approval:
| Deal | Customer | Discount requested | Reason, in a sentence | Pocket price the firm will keep (from page 2's ledgers) | Days to period end | Approved by | Reviewed by (not paid on this period) |
|---|---|---|---|---|---|---|---|
Reading it monthly. Count the discounts beyond guidance by week of the period. A cluster in the last fortnight is the calendar buying the price. Read the reasons: year-end is not a reason.
What this page does not do. It does not design anyone's pay. That design belongs to the firm and its people.
From Chapter 5. The history it reads was fixed on page 1; the conversation is held in days sixty-one to ninety.
Who is in the room. Finance, leading. The commercial lead. An account manager. Someone from credit control or billing. Whoever will calculate, and whoever will check.
The questions, in order.
The one-page agreement.
| Practices measured, and the unit for each | |
| History used, and the ordinary range it shows | |
| Adjustments, each with its record and method | |
| What is not counted | |
| Balancing measures, each with its record | |
| Who calculates; who checks | |
| How and how often the result is shown, and to whom | |
| What happens if a record or system changes | |
| Date agreed | |
| Signed |
Reading the monthly figure. Plot each month against the history's ordinary range, with the balancing measures in the same table. Treat a point as news only when it falls outside the range or several in a row move the same way. One good month is not yet a gain.
From Chapter 6.
| Days | Practice | Who | Time it takes | What it produces | Early sign it is going wrong |
|---|---|---|---|---|---|
| 1–10 | Fix the history; list ledgers; choose customers; list agreements; count the queue (page 1) | Sponsor, convenor, finance | Convenor about 3 hours a week | A dated history and a first-guess page | People asking whether it is about commission |
| 11–30 | Waterfalls and the band (pages 2 to 4) | Finance, account managers | Two hours per waterfall; an hour for the band | Bars with answers; the unexplained spread | Every bar defended |
| 21–45 | The register (pages 8 and 9) | Commercial lead, administrator | About a fortnight of one person | Clauses, dates, decisions | A date passes with no name beside it |
| 31–60 | The scope ledger and the walk (pages 6, 7, 10, 11) | Service manager, engineers; credit control | A minute a line; fifteen minutes a week; two hours to sort and walk | Out-of-scope work seen; one first-desk question | Fewer lines each week; a desk that feels blamed |
| 46–75 | Customer conversations (page 5) | Relationship holders | Fifteen minutes to prepare; the call | Agreed, released and refused changes | A letter before a call |
| 61–90 | Discount record; the measurement conversation (pages 12, 13) | Sales director; finance leads | Two meetings of ninety minutes | A signed page and first monthly figures | Someone in the room goes quiet |
From Chapter 6. The working group and the sponsor, about two hours.
Bring the first-guess page, the waterfalls and band, the register, the scope ledgers, the walk sheets and the signed measurement page.
Ask, in order:
What not to conclude. That the leaks are closed (ninety days show whether practices are kept). That one good month is a gain. That an empty ledger or a quiet register means nothing is leaking.
From Chapter 6. For anyone who wants to begin with less.
The price a firm receives is settled below the invoice, in a dozen places nobody adds up. Most of what leaks there was earned already.
Luminous Prosperity Inc. takes this work into organisations as a gainshare. The flow is opened from the inside, with the people who already hold what the firm knows. The gain is counted from records the firm already keeps, against a measure both sides sign before anyone knows the answer, and the house is paid as a share of what is found: between thirty and ten per cent of it.
The book walks the waterfall, the ledger and the drawer of clauses. Where your own revenue settles, and what recovering it would return, is read from your records and the people who keep them. That begins with a conversation.