Your numbers are real. Your retention is genuinely good. You have been in the market three years and the people who know your work take it seriously. And the round is not moving.
Those are two different systems and you have been treating them as one. Proof is a property of the work. Credit is the output of an institution.
Why these three models
The decision is whether to spend your next quarter improving the work or making it legible. The features that fire are private information a counterparty cannot verify cheaply, a long gap between doing something and being recognised for it, and a reward distribution that is nothing like an average.
Three lenses. Certification produces an equilibrium answer about who converts your evidence into something a stranger can act on. Delay produces a complex answer about the lag and about the damage founders do inside it. The power law produces a random answer about how recognition distributes once it does arrive. The first two disagree usefully: one says buy the verifier, the other says you will be waiting either way, so plan for the wait.
1. The institution in the middle
Start with the cleanest available demonstration that correctness and recognition come apart, and it is not from business.
In November 2002 Grigori Perelman posted a preprint to arXiv and emailed about a dozen mathematicians. The message described a monotonic expression for Ricci flow and said it verified several assertions related to Hamilton’s programme, offering a sketch of an eclectic proof. It did not claim to have proved the Poincaré Conjecture. He said so only when someone asked him directly.1
The proof was correct on the day it was posted. What followed was years of independent teams working through it, a plagiarism controversy in which another group claimed the credit, the Fields Medal declined, and the Clay Institute’s million dollars declined.1 Nothing about the mathematics changed across that period. Everything about the recognition did.
The mechanism is certification. A counterparty who cannot verify your quality cheaply prices you as the average of the pool you appear to be in. What breaks that is a third party whose confirmation is cheaper to trust than your assertion. The founder version is unglamorous: an audited figure, a customer who will take the call, a regulator’s licence, a named investor already on the cap table.
And where verification is available, the pressure to use it compounds. When Los Angeles County put hygiene grade cards in restaurant windows, the cards caused inspection scores to rise, consumer demand to become sensitive to hygiene quality, and foodborne illness hospitalisations to fall.2 The disclosure changed what firms did, not just what buyers believed. That is what a working certification layer looks like from the inside.
Certification, the conversion lens
- Assumes: the counterparty cannot verify quality cheaply, and a third party can.
- Fits because: the work is good and the market is pricing you as the average.
- Breaks when: no credible verifier exists in your market, which is common and is a different problem entirely.
- Evidence: grade A. Robust theory with a clean field demonstration.
- Counteracts: the belief that better work will eventually speak for itself.
- May reinforce: collecting credentials as a substitute for the work.
2. The lag, and what you do inside it
The second lens is the one that explains the specific damage, because the gap is not neutral time.
Recognition arrives long after the work. That delay is structural: verification takes as long as it takes, and the people doing it are not on your schedule. The failure is not the delay itself. It is what founders do while correcting against it.
The pattern is the standard oscillator, and the difficulty is not carelessness: reasoning about delay and accumulation defeats highly educated adults in controlled experiments, and the failure is not attributable to graph literacy, contextual knowledge, motivation or cognitive capacity.3 You act on a signal that is stale, at full strength, and overshoot. Six months of silence reads as a verdict on the product, so you pivot. The pivot lands just as recognition of the previous work arrives, which now describes something you no longer do. So you correct again, in the other direction, against another stale signal.
The two available repairs are the same two available in any delayed system. Shorten the delay, which is what buying certification does. Or damp your response, which means deciding in advance how long silence must run before it counts as information.
Correcting at full strength against delayed information guarantees overshoot. That is not a metaphor borrowed from engineering. It is the same structure, and it produces the same trajectory.
Delay and oscillation, the lag lens
- Assumes: the response to your action arrives after a lag, and you are correcting against the observed signal.
- Fits because: the gap between doing the work and being recognised is long and you are acting inside it.
- Breaks when: the silence genuinely is a verdict, in which case damping your response is exactly the wrong move.
- Evidence: grade A. Structural, and the human failure to reason about delays is well replicated.
- Counteracts: reading a quiet quarter as a result.
- May reinforce: patience with something that is genuinely not working.
3. How recognition distributes when it comes
The third lens governs what the arrival is actually worth, and it is the one that changes where you aim.
Recognition does not distribute normally. A small number of names absorb most of the attention in any category, and the tail is long and thin. That is why the median well-regarded company in your sector is far less well known than the mean would suggest, and why aiming at “being known” produces almost nothing while aiming at “being the named example of one specific thing” occasionally produces a lot.
Two consequences follow. First, a small increase in general visibility is worth close to nothing, because you are moving within the thin part of the distribution. Second, becoming the reference case for a narrow category is a different bet with a different payoff shape, and it is available to companies that are not the largest. Aim to be the named example of one thing, not to be generally known.
The honest caveat belongs in the body. Concentration is a real and observable feature of attention. The formal claim that such distributions follow a fitted power law with a specific exponent is much weaker than the popular version suggests, and this article uses the shape rather than the parameter.
Power-law distribution, the payoff-shape lens
- Assumes: attention concentrates, so the mean is a poor guide to the typical case.
- Fits because: you are deciding how much general visibility is worth buying.
- Breaks when: the market is genuinely fragmented and locally decided, where the concentration never forms.
- Evidence: grade B plus for the concentration. Barred for any specific fitted exponent.
- Counteracts: planning against an average outcome in a domain with no typical case.
- May reinforce: lottery-shaped strategy dressed up as positioning.
The levers, cheapest first
- Name the verifier in your market. Who does a stranger already trust about companies like yours? An auditor, a regulator, a specific investor, a named customer. If you cannot name one, that is the finding.
- Convert one assertion into a confirmable fact. Not all of them. One. A figure your accountant will state, a customer who will take a call.
- Decide the silence threshold in advance. How many weeks of no response constitutes information? Write the number down now, while you are calm, and stop reading the weeks before it.
- Pick one narrow category to be the example of. Broad visibility spends against a distribution that will not pay it back.
- Stop improving the work as a response to silence. If the work was already good, more of it does not address the constraint, and it is the most comfortable thing to do.
What to do this month
Do now, sized at one afternoon, effect visible in weeks. Write down the single strongest claim you make about the business, then write down who could confirm it without asking you. Reversible, free, and dominant across every scenario about why the round is slow.
Hedge, where the premium is the whole loss, live before the next cycle. Get one third-party confirmation attached to that claim. If credibility was never the constraint you have spent an afternoon collecting a reference, and that is the entire downside.
Defer and trigger, size fixed now. Do not rebuild your positioning around a narrow category this quarter. Pre-commit the trigger: if the silence threshold you just wrote down is crossed with the certification already in place, then the constraint was not legibility and the strategy gets reopened. Fix the threshold now, because a threshold set during a slow month is set under pressure.
Watch the arrivals as much as the dates. Certification lands in weeks. Recognition lands after that, on someone else’s schedule, and the interval between them is exactly where the overshoot happens.
What usually happens next
Run the break test first. Has a rule changed, has an actor entered or left, has a measurement become a target? If a new certification body has entered your market in the last year, the historical relationship between proof and credit no longer holds, and the lag may be much shorter than your experience suggests.
If nothing broke, the shape is a long flat period followed by a step. Founders read the flat period as a trend and act on it, which is the specific error described in section two. The step, when it arrives, is usually triggered by one verifiable event rather than by accumulated quality.
There is a second regularity worth expecting. The verifier you route through will be slower and more conservative than you are, because their incentive is to avoid confirming something false rather than to help you raise. Founders read that caution as indifference and often abandon the process weeks before it would have paid, which is the same overshoot described above wearing different clothes.
Subtract the counterfactual before crediting a campaign. Recognition that arrived in the quarter you also shipped a large named customer probably tracked the customer. The customer was the certification.
What this ensemble cannot see
All three lenses assume the work is actually good. None of them can tell you whether it is.
That is the load-bearing assumption and this framework will not test it. Every argument here is equally available to a founder whose product is mediocre and who has concluded the market simply cannot see it. The certification lever is genuinely useful and it is also the most comfortable possible explanation for a slow round, which should make you suspicious of how readily it fits.
There is also a limit on the Perelman case. It is one instance, drawn from a field with unusually clean standards of proof, and it is being used illustratively rather than as evidence. Mathematics has an adjudication mechanism that markets do not. In your market there is often no fact of the matter to be eventually confirmed.
And one property none of these models contains: routing through a verifier changes what you build. Once an auditor, a regulator or a marquee investor is the channel through which your work becomes legible, their criteria start shaping your roadmap, and that influence does not appear anywhere in this analysis.
The one action that survives the ignorance: this week, take your strongest claim to one person who has no stake in your company and ask what would make them believe it. If the answer is something you could obtain in a month, the constraint was legibility. If the answer is that they would need to see the product work for a year, the constraint was never credit, and no certification will shortcut it.
Who has to move
The person who needs this is the founder, and the founder is the one whose instinct under silence is to improve the product, because that is the part they control and enjoy. The cheapest first test is the two-line exercise: strongest claim, and who could confirm it. If the second line is blank, you have found the constraint, and it is not the work.
Sources and notes
- Masha Gessen, Perfect Rigor: A Genius and the Mathematical Breakthrough of the Century, Houghton Mifflin Harcourt, 2009. The November 2002 arXiv posting and the accompanying email, including its description of a monotonic expression for Ricci flow and a sketch of an eclectic proof of the geometrization conjecture, are quoted in chapter 9. The prologue records that Perelman did not publish in a refereed journal, did not agree to vet the explications written by others, refused numerous job offers, and refused the Fields Medal, alongside the plagiarism controversy in which another group claimed credit. Cited here illustratively; mathematics has an adjudication mechanism that markets lack.
- Ginger Zhe Jin and Phillip Leslie, The Effect of Information on Product Quality: Evidence from Restaurant Hygiene Grade Cards, Quarterly Journal of Economics 118(2), 2003. Open-access copy: https://drum.lib.umd.edu/bitstreams/1fce72fc-166b-49ba-8167-fc69cd0b13a6/download. The abstract states that the grade cards cause restaurant health inspection scores to increase, consumer demand to become sensitive to changes in restaurants’ hygiene quality, and the number of foodborne illness hospitalisations to decrease.
- Matthew A. Cronin, Cleotilde Gonzalez and John D. Sterman, Why don’t well-educated adults understand accumulation? A challenge to researchers, educators, and citizens, Organizational Behavior and Human Decision Processes 108(1), 2009. Author copy: https://www.mit.edu/~jsterman/CroninGonzalezSterman061210.pdf. Used in section 2 for the finding that difficulty reasoning about accumulation and delay is not attributable to graph literacy, contextual knowledge, motivation or cognitive capacity, which is why the overshoot described there should be expected rather than treated as carelessness.
A note on a claim this article deliberately does not make. It would be easy to write that attention follows a power law with some exponent, and the popular literature does. Testing across nearly a thousand real networks found scale-free structure to be rare rather than universal, so the concentration is used here as a shape and never as a fitted parameter. The strategic conclusion, that broad visibility spends against a distribution with no typical case, does not need the exponent.
Joshua Agonya Pi’Rwot, Founder.