The Founding 25:Why Your First Partner Cohort Should Be Embarrassingly Small
Your first partners are not a sales channel — they are the founding team of an ecosystem. The case for capping cohort one at 25, hand-picking every member, pricing the free year without poisoning the paid one, and launching oversubscribed instead of racing to 100.
Ask a founder who has just decided to license their methodology what success looks like, and you will usually get a number back: a hundred certified partners. Two hundred. More. The instinct makes sense on the surface — more partners should mean more reach, more revenue, more proof that the model works.
The research says the instinct is backwards. From Andrew Chen's work on cold-start dynamics to Alex Moazed's study of platform ecosystems to Daniel Priestley's demand frameworks, the literature converges on a counterintuitive prescription: cap your first cohort somewhere between 10 and 25, and treat that cap as a strategic weapon rather than a budget constraint.
The ceiling these founders are trying to escape is real enough. Alan Weiss frames it with his Four-Week Vacation test — a business that stops producing the moment you step away is a job dressed up as a company. David Baker, whose research spans 1,340 expertise firms, makes the same point from the buyer's side: a firm that cannot operate without its owner has no acquirer. Scaling through certified partners is the right answer to that ceiling. Scaling through too many of them, too quickly, simply trades one failure mode for another.
What follows is the case for the deliberately small founding cohort: how the door policy works, what smallness actually buys you, who deserves an invitation, how the money has to flow, and how to launch so that the market asks to join you rather than the other way around.
01 — The Door Policy Decides Everything
Three Ways In, and Only One Works for Cohort One
Before you debate cohort size, settle the admission model, because the admission model shapes everything downstream. There are three: invitation-only, open application with selection, and open enrollment.
Invitation-only is the founding-cohort answer, and the community literature is unusually unanimous about it. An invitation gives you total control over who enters. It flatters the person receiving it. And it produces commitment before the program has even started — Seth Godin's observation that the deepest loyalty comes from people who committed before the thing was proven. The strongest articulation comes from Richardson, Huynh, and Sotto: real communities are built with their members, not for them, and co-building is only possible with a group you chose by hand.
Open application with selection has its place — later. It widens the funnel and surfaces talent you would never have found in your own network. The costs: you need real screening infrastructure, and every rejection produces a person who may talk about your program with something other than warmth. Save this model for the growth stage and for entering new geographies.
Open enrollment — anyone with a payment method becomes "certified" — is the one model a methodology business must never adopt, at any stage. A credential that everyone can buy is a credential that signals nothing. Spinks draws the sharper distinction: open enrollment manufactures members, and members consume value. Communities are made of people who create it.
"A credential that everyone can buy is a credential that signals nothing. The day anyone can purchase their way in is the day membership stops being worth anything to the people who earned it."
Hand-selecting the first cohort is more than quality assurance — it is a public statement about what the program means. It tells clients that any partner carrying your mark cleared a bar, and it tells future applicants that the bar exists.
02 — What 25 Buys You That 100 Cannot
Four Compounding Advantages of Staying Small
Different authors reach the small-cohort conclusion by completely different routes. Godin gets there through the "1,000 True Fans" logic of depth over reach. Chen gets there through the Atomic Network — the smallest group that can sustain its own engagement. Priestley gets there through demand mechanics. Moazed gets there through ecosystem research. The destination is identical: a couple of dozen committed partners outperform ten times as many casual ones, and not by a small margin.
Four mechanisms drive the result.
First: at 25, quality is a relationship. At 100, it is a spreadsheet.
With 25 partners you can personally run onboarding, observe early client engagements, read actual deliverables, and have honest one-to-one conversations about what is and is not working. Past roughly that number, quality management goes statistical: you track averages and flag outliers, but you no longer shape any individual partner's practice. The difference between coached quality and audited quality shows up in every client engagement your partners deliver.
Second: the first cohort writes the culture, and the culture does not get rewritten.
In Modern Monopolies, Moazed documents path dependency in platform ecosystems: whatever standard the earliest participants set becomes the permanent center of gravity. Let your first partners shortcut the methodology, discount their way into deals, or treat the certificate as wall decoration, and you have encoded those behaviors into the ecosystem itself — and ecosystems resist cultural upgrades ferociously. Think of cohort one not as your first batch of customers but as the founding team of a community you intend to run for decades.
Third: density, not headcount, generates network value.
Referrals, peer learning, and cohesion all come from interaction frequency, and interaction frequency collapses as groups grow. It is the small-town effect: everyone knowing everyone accelerates trust, and trust is what makes a partner pick up the phone and hand a colleague a client. A 250-person network fragments into cliques almost immediately. A 25-person network stays one conversation.
Fourth: scarcity flips who is selling whom.
Priestley's Oversubscribed thesis is that when capacity is visibly smaller than demand, the power dynamic inverts: candidates stop evaluating you and start auditioning for you. A partner who argued their way into the program behaves differently — permanently — from one you had to chase.
Notice that these four advantages feed each other. Tight quality builds strong culture; strong culture deepens density; density makes the program desirable; desirability creates scarcity, which attracts a stronger applicant pool for cohort two. The flywheel only spins if you decline to go big in year one.
03 — Earning an Invitation
The Filters That Separate Founding Partners from Warm Bodies
Saying no is the discipline that makes the founding cohort work. Every marginal candidate admitted "because we need the numbers" taxes the value of membership for everyone who cleared the bar honestly. The people you reject define the brand as much as the people you admit.
Baker's research gives you the analytical screen. Across his 1,340-firm dataset, five pre-tests predict whether a practitioner has the positioning and depth to thrive inside a partner network:
Competitor count in their niche.
The healthy band is 10 to 200 competitors. Below 10, the market they have chosen is too thin to feed a practice. Above 200, they have not actually specialized.
Twenty insights, on demand.
Baker's "Drop and Give Me 20" — ask the candidate to name, unprepared, 20 lessons their niche has taught them. Real practitioners produce them without effort. Candidates whose expertise is read rather than earned stall around lesson six. This single question filters out more pretenders than any application form.
A market wider than their commute.
A practitioner whose client radius is a 30-minute drive cannot generate the engagement volume that justifies what certification costs them.
Hireable specialists exist.
If they needed to recruit a specialist in their niche tomorrow, could they? A niche in which no specialists exist is a niche the market has not yet validated.
A list you can buy.
Mundane and ruthless: is there a purchasable mailing list of prospects in their segment? If nobody has found it worthwhile to compile one, the segment may exist only in the candidate's pitch deck.
Analytics alone will still admit the wrong people, which is why Michael Port's Red Velvet Rope Policy supplies the second screen — the human one. Score each candidate on the traits that actually predict partnership behavior:
- Energy — after an hour with them, are you more alive or less?
- Coachability — will they run the methodology as designed, or improvise around it?
- Alignment — do they want what you are building, or just the logo for their website?
- Drive — will they go sell, or sit and wait for the directory to feed them?
- Vertical relationships — do they already know buyers in a defined segment?
"Admit only candidates who clear 75% on both the analytical screen and the human one. Yes, this is restrictive. Restrictive is the product."
One more design decision hides in Baker's data: the vertical-versus-horizontal mix. Vertical specialists ("I serve healthcare organizations") versus horizontal specialists ("I do data governance across industries"). His research finds 85% of successful firms choose vertical positioning — prospects are easier to find and industry knowledge compounds — while horizontal positioning trades that for variety and resilience across economic cycles.
Compose the founding cohort with both on purpose. Industry-deep partners plus discipline-deep partners give the network range without sacrificing depth in any single engagement.
04 — A Deal That Holds
The Economics Have to Clear Two Bars Before Launch
Lopsided partnerships do not survive. If one side consistently puts in more than it takes out, the arrangement decays no matter how warm the relationship feels. Model the economics on paper before the first partner signs anything.
The partner's side of the ledger: a license to your methodology, tooling, and diagnostics; the credential and co-branding rights; directory listing and client matching; the cross-referral network; amplification through your podcast, co-authored content, and speaking slots; a peer community with monthly calls and an annual summit; benchmarking data; and sales enablement — playbooks, question libraries, proposal templates.
Your side of the ledger: partners adhere to methodology standards and quality gates; deliver a minimum volume of at least 10 assessments a year; price methodology-branded work on value, never by the hour; show up in the community; contribute thought leadership — case studies, articles, talks; keep client satisfaction above threshold; and report delivery outcomes honestly, including the ugly ones.
Then run the two bars.
Bar one: the 10x partner question.
Can a typical partner credibly expect ecosystem-attributable revenue of ten times their annual fee? A $5,000-per-year certification has to plausibly drive $50,000 or more in leads, referrals, brand leverage, and pricing power the partner would not otherwise have. The pricing-power piece is not hand-waving: Baker's data across 900+ expertise firms shows certified specialists charging 40-100% more than generalists, which makes the credential a fee multiplier rather than a wall ornament. Weiss's Marketing Gravity idea explains the acquisition side — the ecosystem's brand, content, and referral flow slash what it costs a partner to win a client compared to going it alone. If 10x does not pencil, either the fee is too high or the value engine is too weak. Fix one before launch.
Bar two: the self-funding question.
Do total partner fees cover what the ecosystem costs to operate — platform, community management, events, marketing, quality assurance? A network the founder must subsidize forever is not a platform; it is philanthropy with a logo. If the math fails, adjust fees, partner count, or revenue streams until it passes.
"Measure the partner ROI ratio — ecosystem-attributable revenue divided by total membership cost — for every single partner, every year. The target is 10:1 or better, and the number gets reported, not assumed."
Put the whole exchange in a written partner agreement before signature one. Undefined expectations are the leading cause of partner conflict; a contract that spells out what you provide, what they owe, how success is scored, and what happens when standards slip turns assumptions into terms.
05 — Free, Paid, or Both
Pricing Year One Without Poisoning Year Two
Should founding partners pay? Few design questions in partner programs generate more argument, and the literature genuinely splits.
The generosity camp has serious backing. Godin: leaders give before they take. Spinks: social norms bind harder than market norms — and generosity is how social norms get established. Port's counsel is to give too much, then give more. And Chen's cold-start research treats subsidizing the supply side of a young network as standard platform practice, not charity.
The charge-from-day-one camp is just as credible. Baker holds that if nobody rejects your proposals on price, you are underpriced — a third of prospects should walk. Blair Enns is categorical about giving expertise away: "Under no circumstances will we part with our thinking without appropriate compensation." Hermann Simon, meanwhile, names the psychological trap — anchor a relationship at zero and every future price reads as a markup on free.
Weiss points to the synthesis, and it dissolves the dilemma.
Give the founding cohort the free (or heavily subsidized) first year if you can fund it — but never let the price disappear. Invoice at full value and apply a 100% founding grant. The real number appears on every invoice, every renewal notice, every communication, with the grant shown as a line item against it:
"Certification value: $5,000. Founding Partner Grant: -$5,000. Amount due: $0."
Twelve months of seeing $5,000 on paper means the year-two transition lands as the expiry of a grant, not the imposition of a fee. The anchor was set at the true price all along. Framed this way, the free year stops being a discount and becomes what Priestley calls the Remarkable Budget — marketing spend with a higher return than any campaign, because 25 partners who got extraordinary value for nothing become 25 recruiters for cohort two.
"Simon's warning stands: discount a premium position and you have destroyed the position. A free year only works if the value delivered during it makes the paid year feel like the obvious bargain."
Which makes year-one measurement non-negotiable. Track and report each partner's return as the year runs — revenue generated, clients won, referrals received, development completed — so that when the renewal invoice arrives, the partner reads it next to their own numbers and concludes it is the best money they spend, rather than a bill for something that used to be free.
06 — Launch So the Market Comes to You
The Five-Phase Oversubscribed Sequence
Everything above assumes you have more qualified candidates than seats. That condition does not happen by accident — it is engineered, and Priestley's Oversubscribed sequence is the engineering manual. Five phases, each one creating the pressure that powers the next.
Phase one: warm the market before you announce anything.
Publish. Speak. Run free diagnostics. Host webinars. Log every expression of interest — a newsletter signup, a completed assessment, a direct inquiry — as a soft signal. The target is 100x your seat count: 2,500 soft signals for a 25-seat founding cohort.
Phase two: convert interest into intent.
Move soft signals to hard ones with a concrete announcement: a founding cohort, 25 seats, applications opening on a named date — does the person want to be told when the door opens? Count application starts, completed applications, and booked discovery calls. Target 5x seats: 125 hard signals for 25 places.
Phase three: select, never accept.
Open applications only once hard signals exceed capacity — and then choose. Selection itself broadcasts demand: being able to say that 87 people applied for 25 places tells the market more about the program than any sales page could. Nothing here is manufactured; scarcity is just what happens when something worth joining has a fixed door.
Phase four: over-deliver to the people you chose.
Competent is not the bar. The founding cohort's experience has to be remarkable, because their results, testimonials, and word-of-mouth are the entire marketing plan for cohort two. Money spent on cohort one's success keeps paying out for years.
Phase five: make the results loud.
Celebrate founding-partner outcomes in public. This is demand generation, not vanity — once the market watches real partners win, the question it asks changes from whether the program is legitimate to how to get a seat.
"Watch one ratio: qualified applicants per available seat. Under 2:1, the positioning is too weak. Over 5:1, raise the bar or raise the price. Live between 3x and 5x."
And honor the rule most founders miss: the founding cohort should absorb no more than 20% of the demand you collected. Accept 25 of 125 serious prospects and the program is born oversubscribed — with the other 100 forming a ready-made waiting list for cohort two.
Capping cohort one at 25 is not modesty. It is precision. Depth instead of breadth, evidence instead of optimism, a founding team instead of a customer list. The partner networks that compound for decades are built by founders who refused to scale before the foundation could carry the weight.