The Manila Folder: Replace the Funnel with a File the Customer Owns

When a lead fills out a self-discovery form, the ordinary machine already knows what to do. Thank-you email. CRM entry. Nurture sequence next week. Stage two. Probability thirty percent. Sequence B.

That machine is competent at organisation-facing bookkeeping. It is almost useless at the thing the person actually asked for: help working out what comes next.

There is another object. It looks old-fashioned until you notice what it holds. Conversations. Corrected understanding. A responsibility picture. A day they could picture and reject. A rejected option with the reason still attached. A page prepared for a conversation with their partner. The whole journey — including the states you superseded.

Call it a manila folder. Run the relationship as that file, owned by them, and the funnel starts to look like the wrong metaphor for human service.

Our job every time we interact is to be valuable enough to earn the next interaction.

The object you would show them

I first met the manila folder as an accounting cadet. The senior handed it over from the compactors, sat you down, and told you what made this client special. You were being trusted. The folder sat on your desk. The computer was not managing the relationship; it was barely more than a calculator. The understanding lived in the file and in the people who had carried it.

As I argued in Software in the Middle, that object was an operating interface — evidence, interpretation, obligation kept present. This piece is about what happens when the same object turns to face the customer.

In 1993 two things were born in the same year and then forked. Don Peppers and Martha Rogers published The One to One Future, arguing that interactive technology would let companies treat different customers differently and compete for share of customer rather than only share of market.¹² The same year, Siebel Systems was founded — first known mainly for sales force automation, later expanding into the broader CRM market.³

CRM inherited the organisation-facing half: forecasts, stages, visibility. The customer-facing half — a relationship file the person could read and correct — had no technology able to hold it at scale. Graduate CRM textbooks clustered in the early 2000s; academic codification lagged the fork by roughly a decade.⁴

A clean test still separates the two halves:

Consumers already expect personalised interactions, and most get frustrated when they do not receive them.⁵ The usual response is personalised wording around the same sequence. That is not a different object. It is the same funnel with better copy.

Thickness, not stage

A stage is a position on our path. Thickness is work done on theirs.

Thirty pages of accumulated collaboration is visible. It is also an honest moat: people stay because the shared work is real, not because of lock-in. And the richest pages are often the ones you would have deleted in a CRM — the superseded ones. “Liked the business, rejected the role, because of staff management” is worth more than the current summary. A state snapshot deletes exactly that.

Walk the same lead two ways across three interactions and the difference stops being philosophical.

White-glove service is not maximum contact. It is relevant attention, appropriate preparation and continuity at the person’s chosen pace — and the next interaction may be a pause.

A page for one, and a meeting that is already the service

Web pages used to be published to thousands. Now a page can be a piece of communication made for one person at one moment — closer to an excellent adviser laying carefully prepared material on the table than a customer logging into a dashboard.

The form follows the question. One conversation needs a visual story. Another needs a comparison. Another needs a day-in-the-life or a responsibility picture with their part in the middle and agents around the edges. Show recognisable work, not department vocabulary most people do not own. “Would you enjoy meeting someone, understanding their problem and helping them decide?” lands; “Do you want to be in sales?” often does not.

This is recognition rather than recall — Nielsen’s classic usability heuristic: minimise the user’s memory load by making objects, actions and options visible; the user should not have to remember information from one part of the dialogue to another.⁶ Put concrete possibilities in front of people. Let them react. Let “I want to talk about this” open the conversation exactly where it matters. Recognition is easier than recall because cues in the environment support retrieval; interfaces that force free recall create unnecessary cognitive burden.⁷

The meeting is an instance of the service, not a qualification step. Real-time models cannot think deeply mid-call, so do not pretend. Keep a fast lane that listens, reflects and updates the shared page, and a slow lane that researches and returns considered material. A response that feels immediate does not have to have been conceived immediately. Generate when useful; present when helpful.

That separation is no longer only an intuition. Google DeepMind’s Talker–Reasoner architecture explicitly splits a dual-system conversational agent into a Talker that interacts with the user in real time and a Reasoner that performs multi-step planning and reasoning, with delayed synchronicity so the Talker need not wait for the Reasoner before responding.⁸

Asymmetric voice solves the couch problem: the AI speaks and shows; the person taps and types; spoken input is optional for when they are alone. Uploaded documents become contributions to the exploration, not profile completion. A minimum viable company — original, or shaped around them — lets someone try on a possible future. “I like the service, but I wouldn’t want to be the person doing that part” is discovery a questionnaire could never produce. We do not need to understand someone completely before showing them something. Showing them something can help us understand them.

Someone else has to see your value

The folder is the object. The retelling is what fills it — and it is the move people cannot make for themselves.

Michael Polanyi’s durable formulation still stands: we can know more than we can tell.⁹ Expertise becomes invisible to its holder because it became automatic. The Johari window names the same structure from another angle: a blind area of things about ourselves which we do not know, but that others can see more clearly — and which shrinks when others say what they see, in a supportive way, and we are able to hear it.¹⁰

Value is always relative to a market the holder cannot see. A nurse with twenty years in aged care — treat this as a hypothetical illustration — may think she is “just a nurse.” An outside eye can see that she reads a home in ten minutes, knows where the system fails families, and is trusted instantly. Connected to a need she has never heard of — advocacy for elderly people living at home — those same capabilities become a company whose customers feel like her only one, at a reach she could not manage alone.

Part one is three moves: inventory (collect), the outside eye (say which parts are rare), the retelling (rewrite their story so it explains them better than they could). The target testimonial is simple: “Nobody has ever described me like that, and it was true.”

The failure mode is the horoscope. Bertram Forer’s 1949 classroom demonstration still explains why: students given identical generalised personality sketches — copied from a newsstand astrology column — rated them as highly accurate personal descriptions.¹¹ The Barnum effect is the tendency to judge vague, general statements as accurate, unique descriptions of oneself; statements with “universal validity” apply equally to nearly everyone and do not meaningfully distinguish anyone.¹²

Ask for stories, not self-description. “List your skills” gets job titles. “Tell me about a time it was going wrong and you fixed it” gets at tacit knowledge. For an archive: “What did people keep coming to you for?”

A label is a retelling in one word. Wrong labels do work too. I was pushed toward “the IT person.” I rebelled and claimed “chief scientist.” Finding that word unlocked how I saw my own value. “That isn’t me” is easier to say than “this is me,” and it points the same way. Offer candidate labels, some slightly off, and learn from the push-back.

The folder changes hands

Two versions of the folder exist in the imagination. In the first, it is the company’s file on a person — the account team’s folder, kept on our side of the desk. In the second, the person carries it. It sits under their arm. It is packed. It belongs to them.

The second version is better. Anything you can walk out the door with is obviously yours, which settles ownership without a privacy lecture. The owned, portable folder is the data-protection answer. Draw the before-and-after carefully: same person, same posture, same dignity — the only difference is the packed folder under the arm. You are not saying they were less of a person before. You are saying their value existed the whole time and has now been written down.

That single object can run through a longer journey: fill it, carry it to find a fit, open it as a founding file. Three honest exits stop the first gift being a funnel in disguise — pilot a company; lend your know-how to someone else’s; walk away with a better story about yourself. Saying out loud that all three are fine is what makes the help real.

The writing is the crucial part — and nobody does it. I have twenty-eight years of my own exhaust and, by my own account, have only marginally written it down. I have told partners and friendly outsiders how important the writing is. Intelligent people, told directly, still do not write their own folder. It is a chore with no immediate payoff, and you cannot see your own value clearly enough to know which things deserve writing down. That is market validation. The onlooker does the writing. The person talks and corrects.

Why free is strategy

Make the first filling of the folder free, and three things compound at once. Trust: the offer has no hook. Referrals: people tell friends about experiences that changed how they see themselves. Category shape: relative to coaches who charge for conversation and write little down, recruiters who extract a story for their own use, and accelerators that skip the step — the combination of outside eye, written account, owned by the person, and freedom to leave is a create cell in Kim and Mauborgne’s Eliminate-Reduce-Raise-Create grid: factors the industry has never offered, used to lift buyer value and create new demand.¹³ That is a strategic profile, not a census of every competitor.

Every folder teaches the harness what makes a retelling land. Every folder adds to the supply side. Every folder feeds a de-identified population view that is more interesting as shapes than as a pipeline count: people stuck at “can’t picture my day”; deep domain expertise with no wish to manage staff. Person A’s knowledge, person B’s appetite for customer work, idea C that needs both. The sales meeting becomes: which journeys moved because of something we gave them, where are people stuck, which combinations are forming.

Free only scales once the unattended version is good enough. Hand-make the first folders. Cap the weekly number because you do them properly. Track the three exits without punishing grateful walk-aways. And avoid the three betrayals: a folder that flatters; a walk-away that is reluctant or a folder that is really a login; a folder that reads as machine-made.

Two crafts sit side by side. Retelling the idea someone brought — finding the kernel in the noise — is a practised craft; a founder who gives every lead an hour of his time already does it. Retelling the person has no exemplar yet and must be hand-made first. An idea someone brings is still evidence about them: “What have you seen that made you want to fix this?”

Bring clinical expertise in for the care floor — safe, true, non-manipulative, consented; not therapy; a former employer’s documents are not the person’s judgement — and keep ownership of the warmth ceiling and of the folder’s second reader: the agents that will build from it.

What to do on Monday

Pick one live lead. Do not open the CRM stage view first. Open a blank page and write what you actually understand — what they enjoy, what they bring, what they want to avoid — in language you would show them. Put one concrete possibility in front of them. Leave room for correction. When they say something that changes the picture, update the page in front of them. At the end, ask whether the next useful thing is another conversation, a partner brief, a pause, or a walk-away file they can keep.

Then ask the only test that matters: would you show them the record you just made? If not, you are still running a funnel. If yes — and if what you wrote could be wrong — you have started a folder.

SongbirdX is one specimen of this doctrine, not the whole subject. Strip the venture studio away and the folder, the earn-next rule, the thickness metric, the retelling and its horoscope guard still stand. Anywhere the first interaction could be an act of understanding rather than qualification — professional-services onboarding, recruiting, wealth and health advice, B2B account management — the same object applies.

The relationship was never a stage. It was always a file. We just forgot to let the customer hold it — and we never did the writing that would make their value visible, even to them.