A woman in a cream slip dress lifts her face toward the light falling across a pale wall

B-Corps · The marketing line

Marketing that begins with who is reading

A research room where an AI reads a book, a product or an audience through the lenses of human development, watches the field, and writes from what it found.

Begin reading
A woman stands with her eyes closed in the shade of a palm garden
A woman stands with her eyes closed in the shade of a palm garden

What it is

A research room that asks who the reader is first

Most marketing begins with the thing being sold and works outward to whoever might buy it. Developmental Marketing begins at the other end. It asks what stage of growth a reader is living in, what they are reaching for, what frightens them and which words they already use, and only then turns to the cover, the listing, the post and the email. It is a research room for a house that publishes books, builds products and advises clients, and it keeps everything it learns.

At the centre is a project. A project can be a book, a product, a person, an audience, a market segment or a single piece of communication, and a person gives it the material: pasted text, or one uploaded file. An AI then reads that material through a long set of developmental lenses, from the colours of Spiral Dynamics and the stages Robert Kegan described to the six lines of development and the shadow a piece of work carries without meaning to. The assessment comes back as written analysis, not as a single score, and it is kept on the project for as long as the project lives.

The assessment comes back as written analysis, not as a single score, and it is kept on the project for as long as the project lives.

Around that centre sits the rest of the room. There is a brand intelligence agent that answers at length and keeps its conversations. There is web research, asked one question at a time or sent out in batches, with every result saved as an insight. There are customer profiles drawn from that research, client records for the companies a person advises, and a watch on competitors that fetches their pages and records what changed. There is a studio that writes posts, email sequences and headlines from everything the room holds, and that refines its voice from the ratings a person gives.

It does this with a language model, and the page says so plainly because the app does. Nearly every request the app makes to the model carries the house's own canon document beside the material, so the model reads the house's developmental framework each time rather than inventing one. Where a request needs the present state of the world, for research, for a market report, for a book's listing, the model is given live web context; where it needs only the material in hand, it is not.

What the room does not do is pretend the numbers are measurements. A stage, a range of lines scored from one to five, a confidence of low, medium or high: these are the model's reading of the text, set down so a person can argue with them, and every one of them can be corrected by hand in the assessment editor. That is the honest shape of the work. A reading is offered, a person who knows the book or the market disagrees with part of it, and the record keeps the better of the two.

A woman reaches up toward a high window in a stone room
A woman reaches up toward a high window in a stone room

A day with it

An editor at a small press, one Thursday in spring

The editor of a small press opens her mail a little after seven, and the morning brief is already there. It came from the app at seven, New York time, as it does every day: one product worth building, a slogan for it, the developmental stage of the people it is for, and a forecast for the next ninety days. She reads it over coffee the way some people read the weather. She does not build an app every morning. She reads it because it tells her, in one page, what the public conversation was doing overnight.

At her desk she opens the dashboard and sees the counts that matter to her: the projects in the house, how many have been analysed, how many are still drafts, and the five newest insights from the research. A manuscript arrived last week from an author the press has worked with before, a book about grief and the body. She starts a new assessment, chooses book, uploads the file, and the app reads the title and a two-sentence description out of it for her, which she corrects by one word.

She runs the analysis and makes tea while it works. When it returns, the project holds a centre stage and a range, a growing edge, a Kegan stage, six lines of development each scored and explained, and a section on shadow material that she reads twice because it names something about the book she had felt and not said. Below it is the marketing the app writes only for books: a direction for the cover with a palette and type, three titles with subtitles and the reason for each, and an Amazon listing with its keywords and categories. She keeps two of the three titles and strikes one.

Late in the morning she turns to the field. A press she admires changed its catalogue page last month, and she has been meaning to watch it. She adds its address to the competitor watch and checks it; the app fetches the page, an AI pulls out the products, the prices, the headlines and the one sentence that says how the press positions itself, and that becomes the snapshot. From now on the app checks it every twelve hours and records what was added, what was removed and what was reworded.

After lunch she sends five questions to bulk research at standard depth, the market for books on grief, what bereavement groups read, how somatic practice is being written about this year, and two more, and watches each line turn from pending to done. From the clearest of the five insights she asks for a customer profile. It comes back with an age range and an occupation, but what she copies into her notebook is the list of phrases readers use about loss, set down word for word, and the places online where they gather.

In the content studio she asks for three social posts and an email sequence for the book, reads them, rates the best a five and the weakest a two, and notes why. Before she leaves she asks the brand intelligence agent for a positioning report for the author, reads it, and downloads it as a Word document to send on Monday. On Friday morning the weekly synthesis will be waiting in the research hub, written from everything the room gathered this week.

A woman holds a standing split against a pale wall, her face hidden by her leg
A woman holds a standing split against a pale wall, her face hidden by her leg

The assessment

A book, a product or an audience, read through many lenses

The assessment is the heart of the room, and it runs as a short sequence of model calls when a person presses Run Analysis. The first call is the structured reading. It returns a centre stage on the Spiral Dynamics colours with a low and a high end, a growing edge, a reading of cognitive complexity, a Kegan stage, six developmental lines scored one to five, the quadrants of integral theory, the states and types a piece of work invites, and separate readings for field intelligence, psychospiritual development, the body's sense of time, and the capacity to hold a group while something new emerges.

The same call writes the parts a marketer reaches for first. There is a summary and a communication strategy, a note on product design, five to seven sample messages written in the register the reading suggests, and three to five warning flags, the ways a message could misfire with exactly this reader. There is a section on shadow material, its indicators and the protective habits around it, and sections on life circumstances and learning preferences: pace, structure, whether the reader learns in company or alone, whether they want theory or practice first. A confidence level closes it.

The reading is offered as a draft from a careful colleague, and the record keeps whatever the person decides.

For a book, a second call follows with live web context, and it is written for the shelf. It proposes a cover direction with imagery, a colour palette given as exact values, type and visual style, each with its reason. It offers several titles and subtitles, each with the appeal it is aiming at. It writes an Amazon title and subtitle, a long description, seven backend keywords and the categories to list under, and sales copy with a hook and the reader's own difficulties named. For a product, the second call is shorter: positioning, pricing psychology, the order in which features should be built, and a messaging frame.

Two more calls finish the work. One writes actionable insights: next steps in order of priority, resources to read, quick wins and a longer pathway. The other writes a long essay in five parts that turns the whole reading into plain consequences for the real world, what this means in practice, for this audience, this week. Everything is saved on the project, the status becomes analysed, and the report opens as a set of cards a person can read top to bottom, with the lines shown as bars, a line chart or a radar, and each chart saved as an image when wanted. The reading is offered as a draft from a careful colleague, and the record keeps whatever the person decides.

Beside the general assessment sits a book assessment of its own, built on the house's non-dual reading of Spiral Dynamics, in which every stage carries a gift rather than a deficiency to outgrow. It maps a book's stages, digs for the gift in each, positions the book for readers at the Green stage, sketches the ecosystem of things that might be sold around it, weighs the title, gives an overall score and prepares an Amazon package. It reads the customer profiles, insights and knowledge items the room already holds, so a press that has done its research gets an assessment that knows its readers.

Then the Amazon optimiser, which does one thing with care. A person gives a book's title and description and names the developmental transition the intended reader is in the middle of, leaving one stage and entering the next. The AI rewrites the listing for that reader alone and returns it as a listing preview, a set of keywords, a note on the buyer's psychology and the marketing copy. A press with a back catalogue can take each book in turn and see the same pages written for the reader it actually has.

A press that keeps its catalogue in Airtable can evaluate it there. With an Airtable connection, the evaluator lists the bases and tables, reads the book records, and has an AI evaluate each title for marketing, brand and audience; each evaluation is kept as a knowledge item, so the room's memory grows with the catalogue, and an evaluation can be set as a PDF and placed in the person's Google Drive. Because every evaluation lands in the same memory as the research, the next assessment of a new book can read what the room already knows about the books beside it on the shelf.

A woman in a white blouse and a gold cuff touches her face beside a monstera leaf

A reader decides in a few seconds whether a page was written for them. The work of knowing them takes considerably longer.

A woman sinks into a low lunge with one arm raised, eyes closed, as leaf shadows cross the wall
A woman sinks into a low lunge with one arm raised, eyes closed, as leaf shadows cross the wall

Research

The agent, the web, the reader and the competition

The brand intelligence agent is the room's standing analyst. It is instructed as a senior synthesis of a brand strategist, a market researcher, a developmental psychologist and a product designer, and it is told never to be brief when depth is available. A person talks to it in conversations that are kept and listed, so last Tuesday's thread on positioning is still there to return to. It can read the room's projects, client profiles, articles and signals, and it can add projects, profiles, objectives and insights of its own when a conversation produces something worth keeping.

Its answers are written as reports: an executive summary of the key findings, numbered sections, strategic recommendations in order of priority, and a plan for thirty, sixty and ninety days. It works across market research, brand architecture, mapping a product to the stage of its buyer, feature specifications for an app, store listings, competitive intelligence and content strategy, and it will synthesise any pasted notes or transcripts into a structured analysis. Any answer downloads as a PDF, a Word document or Markdown, which means a client receives a document rather than an invitation to a system. A link on the page connects the same agent to WhatsApp.

A customer profile is only as good as the words it keeps, and this one keeps them verbatim.

Web research comes in two sizes. A single question goes to the agent through the web research page, and the answer arrives in the same conversation. A batch is a list of addresses and topics, sent together at quick, standard or deep depth; each is researched by an AI with live web context, saved as an insight, and the batch shows each item moving from pending to running to done, or to failed with the reason, so a person can see what came back and what needs asking again. Every Monday the room repeats this on its own, re-researching the keywords its insights care about most.

All of it lands in the research hub. An insight keeps its headline and its full text together with the address or keyword it came from and the keywords it found. It carries tags and a developmental focus, a priority and a status, and the hub lets a person filter the insights, compare two side by side, chart them by week and export them. From any insight a person can ask for a customer profile. The AI sets down the demographic outline first, an age range and an occupation with income and education. Then it writes what the reader values and believes, what they fear and hope for, and the statements they use about who they are. It ranks the jobs they are hiring a product to do, names the moments that make them buy and the objections that stop them, and records the formats they prefer and where they gather online. A customer profile is only as good as the words it keeps, and this one keeps them verbatim.

Client records sit alongside. A client profile holds the company and its industry, the messaging it uses now, the stage of the audience it wants to reach and a description of that audience, its goals, the measures it is judged by, and the market around it. A person who advises several companies keeps one record for each, filters them by industry or by audience stage, and the commercial reports can set a project against a client's profile when the question is whether the two belong together. The client dashboard counts the assessed projects and the signals held, and shows how readings of stage have moved over time.

The competitor watch is the plainest of the room's instruments, and the most useful on an ordinary morning. A person adds a competitor's name and address and marks it direct, indirect, aspirational or a source of content. The app fetches the page itself, sets aside the scripts and styling, and hands the text to an AI that returns the products, the prices, the headlines, the claims, the call to action and one sentence of positioning. The next check compares the new snapshot with the old: each addition and removal is recorded, a change in pricing is marked as significant, a change in positioning is recorded as it was and as it is now, and the room keeps the most recent history beside the previous snapshot. A scheduled job checks every active competitor every twelve hours.

A woman with curly hair sits with her eyes closed and her palms joined above her head
A woman with curly hair sits with her eyes closed and her palms joined above her head

Writing

What has not been said, and the words to say it

Research is only worth what it lets a person write next, and the room is arranged to answer that question directly. The content gaps page asks an AI to read the stored research beside the articles already written and to name what has not yet been said: the subjects the audience is asking about that nothing on the shelf answers, and the angles a competitor has left open. It is the shortest route in the room from a week of reading to a list of pieces worth writing.

The app also keeps a written editorial plan for its blog, set down in full on its own page rather than generated: the reader it is for, a craftsperson who distrusts corporate speech and buys on resonance, and the categories the blog writes in, with Field Notes as the primary voice, each with its reason, its tone and sample titles. It is a plan of record, the kind an editor pins above the desk, and it sits beside the gap analysis so a person can hold what the research found against what the house intends to say.

The ratings are the point.

The content studio is where writing happens. A person chooses a social post, an email sequence or a blog headline, names the platform and the topic, and the AI writes several pieces, drawing on the room's strategic report, its knowledge items, a persona and the developmental stage being addressed. Each piece arrives as a card with its hook. A person rates it from one to five, adds a note and a quick tag (too formal, too casual, off brand, on point) and saves it. The ratings are the point. When the voice panel is asked to refine the voice, an AI compares the pieces rated four and five with those rated one and two and keeps a refined voice and persona profile among the knowledge items, and the writer reads that profile the next time it writes.

The campaign builder starts from the book itself. A person uploads the manuscript, the app extracts its content from the file, and an AI writes a campaign meant to be run that week: an inventory of the assets the author already has, what the book itself suggests about its readers, and the campaign, laid out so it can be followed rather than admired. It is written for the author who has a book, a website and a phone full of short videos, and no plan connecting them.

The strategy report is the room's longest single document. An AI reads every insight, customer profile, knowledge item, business objective, article and signal the room holds and writes one strategic intelligence report from all of it: brand strategy, market research, the psychology of the ideal customer, and a roadmap for going to market, at a length and finish meant for a client's table. The content studio reads this report when it writes, so the posts and emails a person rates are grounded in the same strategy the client was shown.

A long report kept in the research hub can be opened as an edition: a cover, a table of chapters, each chapter set on its own page with plates, and a player when a recording has been attached to it. It turns a piece of research into something a person reads slowly, the way the house reads its own issues. The strategy room holds a whiteboard on which a person draws connections between directives and goals, titles the session and saves it; a strategic synthesis page lets a person choose among the stored developmental signals and have an AI write a synthesis across them.

A woman leans against a sunlit wall with her arms folded, looking away into the light
A woman leans against a sunlit wall with her arms folded, looking away into the light

Reports, rhythms and connections

What runs on a schedule, and what plugs in

The commercial reports are the room's formal instruments, each written by an AI with live web context from a project's assessment. There are five. Product and consciousness fit asks whether a product speaks to the stage of the people it is meant for. Why they will not buy names the objections and the mismatches that keep a willing reader from paying. Consciousness due diligence reads a company or an offer before a person commits to it. A market emergence forecast looks at where an audience is moving. A conversion audit reads a page or a funnel for the places a reader leaves. A comparative analysis sets several projects against client profiles in one document.

Some of the room's work runs without being asked, on a fixed rhythm. The morning brief runs at seven each day: the app reads twenty public sources, among them the top charts of the App Store, Google Trends and Google News, Hacker News, GitHub's rising projects, the most-read pages of Wikipedia, several communities on Reddit, the World Bank and exchange rates, adds the stored developmental signals and the last seven briefs, and has an AI choose one product and write its forecast, its marketing and a build brief for a coding assistant. The brief is kept in the Product Predictor and emailed to the admin.

Some of the room's work runs without being asked, on a fixed rhythm.

The other rhythms are quieter. Every twelve hours the competitor watch checks each active page. Every Monday the room re-researches its most important keywords and looks over its research for age: anything older than six months is tagged stale and needing a refresh and its priority lowered, never deleted. Every Friday an AI synthesises the week's active research into a strategic intelligence report and files it in the research hub. Report schedules for the commercial reports are kept as records of their own, which projects, which report, weekly, monthly or on completing an analysis, how and to whom, so the house's intended rhythm is written down beside the work.

The library is where finished readings live. It lists every analysed project, with search, a filter by kind and by stage, and a choice of order, and the whole of it exports as one file. Tags order it further: each tag has a category (consciousness, developmental, market, industry or custom), a colour and a description, and a tag's own view lists the projects that carry it. The knowledge base is held in the intelligence hub, where books, apps, courses, articles, research, podcasts and frameworks are kept with their themes, their audience, their stage, the segments they serve, their key concepts and a score from nought to ten for how well the room understands them, and where an AI can be asked for a multi-part analysis across the knowledge, the research and the profiles together.

Other systems connect by key. A person makes an API key with a name, an organisation, a tier and the endpoints it may reach; four endpoints answer to it, one returning a project's assessment, one a comparative analysis, one predictive signals and one the Epoch Lense analysis, and each refuses a key that is missing or inactive, while the signal endpoints also refuse a key that was not allowed their door. The Epoch Lense request sends a market segment and an audience to the house's own forecasting service, has an AI interpret the result against the business objectives the person has set, stores the signal, and posts it to any webhook whose conditions it meets. A documentation page describes each endpoint, and the webhook settings hold each address with its conditions, its retry settings and a count of how often it has fired.

A woman in an open-backed cream dress walks through a pale living room toward the sea

What a competitor changed overnight is rarely announced. It is simply there in the morning, on the page, for anyone who looks.

A woman crouches into a low lunge on a wooden floor, her eyes on her hand
A woman crouches into a low lunge on a wooden floor, her eyes on her hand

How it joins B-Corps

The research beside the voice, the pages and the people

B-Corps is the house's own operating system: nineteen applications installed side by side on one desk that starts each one, opens it, rests it, and every two minutes restarts any that has stopped. Developmental Marketing is the research and brand-intelligence room in its marketing layer, and it keeps its own sign-in and its own records, as each of the apps does today. What follows is how the suite is arranged to be used together; where data passes between apps, it is because a person carried it.

The nearest neighbour is The Luminous Book, the brand studio at the centre of the suite, which reads a business, writes down its voice and publishes in it. A person who keeps the two open side by side finds that they answer each other. The voice The Luminous Book sets down is exactly what a client profile's current messaging and a project's source material want, and the refined voice and persona profile this room keeps from a person's ratings is the kind of evidence a voice should be checked against from time to time.

A person who keeps the two open side by side finds that they answer each other.

SquareDash is the marketing workspace for websites, its pages, articles, search and social, and it has been given the house's whole catalogue of products and The Luminous Book's brand voice to work from. Developmental Marketing is the room that tells it what to write. The content gaps found here, the exact phrases kept in a customer profile and the shifts recorded by the competitor watch are what a person brings to SquareDash when deciding the next article or the next page, and its articles go on as drafts to the house's publication through the bridge the suite already keeps.

For a consultant, the room sits beside the foundation and the assessment tools. B-Corp holds one record of an organisation, its people, teams and impact, and a client profile here holds the same company seen from the market's side, its messaging and the stage of the audience it wants. Resonance Canopy assesses an organisation from within; this room assesses its audience from without, and a person holding both sees an offer from two directions at once. Resonance CRM keeps the sheet of people a firm intends to reach, and the customer language gathered here is what the first letter to each of them should sound like.

For a publisher, the room sits upstream of the press. A book assessed here, with its stage map drawn and its title weighed and its listing rewritten for the reader in transition, goes next to Luminous Bookflow, the publishing house in software, where its page is made. ContentSpark sets one source across every channel, and a campaign written here from the manuscript is a good source to set. The house's reading room reads every app and every book in the suite, so what this room learns about a reader is part of what the house as a whole knows.

A woman rests on a massage table under a wooden roof, wrapped in a towel, gazing into the garden
A woman rests on a massage table under a wooden roof, wrapped in a towel, gazing into the garden

Who it is for

For people who would rather understand the reader than shout

It is for a small press that publishes books on difficult subjects and wants every listing to sound as though it were written by someone who read the book. It is for an author who has a manuscript, a website and an audience that has not yet found them, and who wants a campaign that begins with the reader rather than the algorithm. It is for a consultant who advises several companies and needs, for each, a record of who they are trying to reach and what has been learned about those people over months, not days.

It replaces several things that usually live apart: a brand strategist's notebook of personas, a research assistant's folder of saved articles, a spreadsheet of competitors' prices checked by hand when someone remembers, a separate tool that writes posts with no memory of the strategy behind them, and the long report on a client's market that is written once and never updated. Here they are one room, and each part reads the others. The posts are written from the strategy report; the strategy report is written from the research; the research is kept, aged and refreshed on a schedule.

Most marketing tools sort people by what they bought; this one tries to understand what they are growing toward.

What makes it different is the lens. Most marketing tools sort people by what they bought; this one tries to understand what they are growing toward. It reads a book or an audience through developmental frameworks that are usually kept in a psychology department, Spiral Dynamics, the work of Robert Kegan, the integral map of quadrants and lines, and it reads them in the house's own non-dual register, in which every stage carries a gift. The result is marketing that addresses a reader's growing edge rather than their insecurity, and a warning list of the ways a message could misfire with exactly this person.

It is also different in what it is willing to say about itself. The readings are the model's readings, written from the material a person provides, offered with a confidence and open to correction by hand. A report is a draft from a well-read colleague, and the room is built for the person who reads it with a pencil. That is a narrower claim than many tools make about their scores, and it is the one that holds up when a client asks where a number came from.

How it is offered

Installed for an organisation, and kept

Developmental Marketing is offered to an organisation in three ways: on its own, as part of the marketing line, or inside the whole of B-Corps, where every application on these pages is installed together, set up in the organisation's own voice and with its own records, and kept running for as long as the organisation keeps it. The house carries each application into its suite for good, so what is set up now is still there, and still mended, in the years after.

The fee reflects the work of installing and shaping it, and the size of the organisation it serves. It is set with each organisation, after a conversation about what the application will carry for it.

Arranged in conversationDevelopmental Marketing · B-Corps

A woman in a white halter dress tips her head back, eyes closed, one hand at her temple, on a sunlit terrace
A woman in a white halter dress tips her head back, eyes closed, one hand at her temple, on a sunlit terrace

The care it takes

Kept, aged honestly, and never louder than the work

The care begins with sign-in. The room asks a person to sign in before it opens, with an email and password or a Google account, and the morning brief will run only for an administrator or for the scheduler that wakes it at seven. Its API answers only to a key that exists and is active, and its signal endpoints ask as well whether the key was allowed the door it is knocking on. These are ordinary protections, and they are in the code rather than in a promise.

The care continues in how the room remembers. Research is never thrown away because it is old; it is marked stale and lowered in priority, so a person can see what needs asking again and still read what was found before. A competitor's previous snapshot is kept beside the current one, with the most recent history of what changed and when, so a shift in a rival's positioning can be traced rather than merely noticed. A customer's exact phrases are kept word for word, because paraphrase is where a reader's voice is usually lost.

A reading is a beginning, and the room keeps it until a better one arrives.

And the care shows in the house's framework being present in the work. Nearly every request the room sends to the model carries the house's canon beside it, so the readings, the reports and the posts are written from one consistent understanding of human development rather than from whatever the model happens to assume. The voice of the content studio is refined only from what a person has rated, never from anything they did not choose to judge. Reports leave as documents a client can keep, in PDF, Word or Markdown.

What a person takes from a morning here is modest and specific: a reading of a book that names something the editor felt and had not said, three titles to consider, a competitor's new price, the five phrases a grieving reader uses, and a post rated well enough to be the model for the next one. A reading is a beginning, and the room keeps it until a better one arrives.

B-Corps

It begins with a conversation

Request a conversation