Your pricing page says contact us. The reasoning is sound: every deal is different, publishing a number anchors the negotiation low, and your competitors would see it.
All three of those are true about the deals you are in. None of them is about the buyers who priced you in eight seconds and closed the tab.
Why these three models
The decision is whether to publish a price, and for whom. The features that fire are a buyer who cannot evaluate you individually and will substitute something, a population that sorts itself before you have any contact with it, and a revenue base concentrated enough that the aggregate effect is not the relevant one.
Three lenses. Statistical discrimination produces an equilibrium answer about what fills the gap when you say nothing. Self-sorting produces a complex answer about what happens in the market before anyone contacts you. Concentration produces a random answer about whose behaviour actually decides the outcome. The first two argue for publishing and the third is the reason it is not automatic.
1. The gap gets filled, and not by you
An absent price is not treated as unknown. It is treated as an estimate problem, and the estimate comes from whatever the buyer can observe: your category, your competitors’ published numbers, how expensive your website looks, and their last purchase in the same space.
That substitution has a name and it is not irrational. When evaluating an individual case is costly, decision-makers use the average of an observable group, which is a reasonable response to a screening cost rather than a prejudice against you.1 It is also durable for the same reason: you cannot argue someone out of it, because the entire point is that they are not evaluating individual claims.
Which produces an outcome most founders have backwards. If your category’s typical price is above what you charge, hiding is expensive: buyers assume you are out of reach and never start. If your category’s typical price is below yours, hiding is protective in the short run and creates a second problem, because the buyers who do make contact arrived expecting the lower number and the whole conversation is spent moving them off it.
So the first question is not whether to publish. It is what number your buyers currently believe. That is answerable and almost never asked: put it to five people in your target segment who have not bought from you, and ask what they would guess you cost. The spread in those answers is more informative than the average.
Read the spread rather than the midpoint. Five guesses clustered within a narrow band mean the market has a settled view of your category, and your only decision is whether your real number sits inside it. Five guesses running from a tenth of your price to three times it mean the buyers cannot tell what category you are in at all, and no pricing page fixes that, because the number will be read against whichever category each of them assumed. In that case the publication that helps is a sentence naming the alternative you replace, before any figure appears.
The substitute they reach for is usually not a competitor. In East African markets the comparison a buyer makes is frequently against a salary, against what an existing staff member is already doing badly for free, or against the last similar invoice they paid in a different currency two years ago. Ask the five people what they compared you to, not only what they guessed. The comparison sets the number; the number is downstream. A software price judged against a junior salary lands differently the moment you say which of that person’s tasks it removes.
2. They sort themselves before you hear from them
The second lens covers the part of the market you never observe, which is why it is systematically underweighted in this decision.
A market is a population of buyers each applying their own simple rules: too expensive, too cheap to be credible, worth a call, not now. Those rules run whether or not you are in the conversation, and the aggregate pattern that results is not something anyone designed.2 A published price is an input to every one of those rules simultaneously. An absent price is also an input, and its effect is to move a portion of the population into not worth the effort of finding out.
Every enquiry you receive is a survivor of a filter you did not write and cannot see.
The practical consequence is a genuine benefit that the contact-us reasoning ignores entirely: a published price does your qualification before anyone reaches you. Sales time stops being spent on buyers who were never going to pay, and the meetings you take start further along.
The measurable version is a comparison you can actually run. What share of your first calls end because of price. If it is high, you are paying in sales time for the information a published number would have given away for free, and that is a cost with an invoice attached even though it never appears as one.
There is a middle option that gets dismissed too quickly. Publishing a starting price, or a band with the conditions that move it, captures most of the sorting benefit while leaving the negotiation open. The argument against publishing is usually an argument against publishing one fixed number, and that was never the only choice.
The band only sorts if you publish what moves it. A range from one figure to four times that figure with no explanation shifts the guessing rather than ending it, and the buyer who cannot place themselves inside the range behaves exactly like the buyer who saw no price. Name the two or three conditions that decide where a buyer lands: seats, sites, volume, whether you do the migration. Then a reader can put themselves at a point, which is the whole mechanism, and the enquiries that arrive already know roughly what they are agreeing to.
Watch for the sorting working in the direction you did not want. A published floor filters out buyers below it, which is the intended effect, and it also filters out the small buyer who would have grown into a large one over three years. If your best current account started at the bottom of your range, the sorting is removing your own pipeline. The fix is not to hide the price again, it is to publish a smaller entry point deliberately, priced to be bought rather than to be profitable.
3. Whose behaviour actually matters
The first two lenses argue for transparency. The third is the reason this is a real decision rather than an obvious one.
Revenue in most companies is concentrated: a small number of accounts contribute a large share of the total, so the mean account is far larger than the median and the average buyer describes almost nobody.3 The sorting benefit accrues across many small buyers. The anchoring cost lands on a few large ones.
That asymmetry decides the answer, and it decides it differently for different businesses. If your revenue comes from many similar customers, publish: the sorting gain applies to the population that matters and the anchoring risk applies to almost nobody. If your revenue comes from a handful of large negotiated deals, the calculus inverts, because one anchored negotiation can cost more than a year of improved qualification.
The honest version for most early companies is that they are somewhere in between and moving. A company whose first six customers were large and negotiated, and whose growth now depends on a long tail of smaller ones, has a pricing page designed for a business it is leaving.
There is a further asymmetry inside the anchoring worry that is worth naming, because it is what makes the fear larger than the cost. A published number becomes the reference the whole negotiation is measured against, and a reference is much harder to move than an opinion.4 That is real. It is also why the useful publication is a floor or a band rather than a point, since a floor anchors upward and a point anchors both ways. Publish the number you will not go below, not the number you expect to get.
The concentration test has a trap in it, which is that it is measured backwards. Last year’s revenue tells you which buyers you served, not which ones you could reach. A company holding four large accounts because the founder personally knew four large buyers is concentrated by history rather than by economics, and reading that as a property of the market keeps the pricing page pointed at a business that cannot repeat itself. Run the split twice: once on revenue, once on your last fifty enquiries. If revenue is concentrated and enquiries are not, the volume segment already exists and is being turned away at the pricing page.
The resolution is segmentation rather than a single answer: publish for the segment where volume lives, and keep enterprise as a stated conversation. That is not a compromise, it is the arrangement both models actually recommend once you notice they are describing different buyers.
What the three say together
- Ask five non-customers in your segment what they would guess you cost. The spread is the finding.
- Count what share of first calls end on price. That is what the absent number is costing you in sales time.
- Check where your revenue is concentrated. Many similar buyers argues for publishing; a few negotiated deals argues against.
- Consider the middle: a starting price or a band with the conditions that move it, published for the volume segment only.
Where they disagree
The sorting argument and the concentration argument are in direct conflict, and it is not resolvable in the abstract.
Sorting says publish, because the buyers you never hear from are a larger and more expensive loss than the negotiating room you preserve, and because the loss is invisible which is why it never enters the discussion. Concentration says that your outcome is decided by a few large deals, that a published number weakens your position in every one of them, and that optimising for a population which contributes a minority of revenue is optimising the wrong thing.
Both are correct about their own population, so the resolution is the revenue split rather than a principle. Compute what share of last year’s revenue came from deals above the price you would publish. If it is most of it, the concentration argument governs and you should stay quiet or publish only a floor. If it is a minority and falling, the sorting argument governs and the pricing page is already out of date. That is a number you have, and it converts a philosophical argument into arithmetic.
What none of them contain
None of the three prices what a published number does to your existing customers. They will see it, they will compare it to what they pay, and any customer paying more than the published price now has an agenda item for the renewal.
None of them handles the market where prices are not comparable because the scope is not comparable. In genuine project work, a number without a scope is misleading rather than transparent, and the useful publication is a worked example rather than a price.
None of them accounts for what publishing does to your own discipline. A number in public has to be defended internally, which means discounts become exceptions that someone has to approve rather than adjustments nobody records. That is a benefit and it is also a constraint, and companies that publish usually discover their real discounting behaviour for the first time in the quarter after they do it.
And one property the ensemble will not produce: publishing changes what you can charge later. A number in public becomes a reference point for your own future decisions as well as your buyers’, and companies find it harder to raise a published price than an unpublished one, even by the same amount. That is a real cost of transparency and it arrives years after the decision.
The one action that survives the ignorance: work out what share of last year’s revenue came from deals larger than the price you would consider publishing. That single number tells you which of the two arguments governs your business, and it is in your accounts already.
Who has to move
The founder, because sales will reliably argue against publishing and their reason is real: it makes their next negotiation harder. What sales cannot see is the buyer who never called, so the cost they are avoiding is visible to them and the cost they are causing is not. The cheapest first test is the five guesses from non-customers, which takes a morning and frequently reveals that the market has already assigned you a price and it is not the one you charge.
Sources and notes
- Jeffrey Carpenter and Andrea Robbett, Game Theory and Behavior, MIT Press. Statistical discrimination, in which decision-makers substitute the average of an observable group when individual assessment is costly, is developed alongside signalling and screening. Used in section 1, including the point that it is a response to screening cost rather than to preference, which is why better claims do not dislodge it and why the remedy is to supply the information rather than to argue about it.
- Stefan Thurner, Rudolf Hanel and Peter Klimek, Introduction to the Theory of Complex Systems, Oxford University Press. Agent-based treatments, in which a population of individuals applying simple local rules produces an aggregate pattern nobody designed, are developed in the chapters on complex adaptive systems. Used in section 2 for the claim that market-level sorting happens without your participation and is invisible from inside the funnel. The model is used qualitatively; no simulation is claimed.
- Albert-László Barabási, Network Science, Cambridge University Press, on heavy-tailed distributions and the divergence between mean and median where a small number of cases account for a large share of the total. Used in section 3 for the general property only. On why the stronger label is avoided, see Anna D. Broido and Aaron Clauset, Scale-free networks are rare, Nature Communications 10, 2019, article 1017, https://www.nature.com/articles/s41467-019-08746-5, reporting that only a small minority of nearly one thousand network datasets show the strongest evidence of scale-free structure.
- Sanjit S. Dhami, The Foundations of Behavioral Economic Analysis, Oxford University Press. Reference dependence, under which a figure is evaluated against a comparison point rather than in absolute terms, underpins both the anchoring concern in section 3 and the observation in the blind-spot section that a published price becomes a reference for your own later decisions. Direction only; no coefficient is quoted, since the magnitudes circulating in business writing are contested.
A note on a number this article does not give. There is no threshold share of revenue above which you should stop publishing. The trade depends on your sales cost per call and on how much negotiating room you actually use, both of which are specific to you. What transfers is the comparison, and both of its inputs are already in your own records.
Joshua Agonya Pi’Rwot, Founder.