Manifesto

Why the core is free

Every pricing page in this category is an argument about what software is worth. This is ours, written down so you can hold us to it.

Somewhere in your HR software, there is a report about your own org chart that you cannot run, because you bought the tier with the org chart and not the tier with the report about it.

Nobody designed that outcome on purpose. It emerged from a pricing meeting. HR software began as a database of people, genuinely hard to build in 1995 and genuinely expensive to sell. When the building got easier, the price needed defending, and the industry found its answer: slice one product into Basic, Professional and Enterprise, and let the feature list do the negotiating. The product stopped being the thing you sell and became the thing you withhold.

Twenty years on, the pattern is so normal that we forget how strange it is. The approval workflow exists in the code you already run, but it costs more per person per month to switch on. The audit log is being written either way, storage is not the question, but reading it is Enterprise. Single sign-on, a security feature, is priced as a luxury, which is why there is a website dedicated to naming the practice. And the price of all of it is on no public page, because the price depends on how rich you look on the call.

We ran hiring and HR for other companies for years. We sat on the buying side of those calls, over and over, and watched founders pay for the same trick in every deal: the thing you need next quarter is always one tier up. Feature-gating is a tax on growing, levied at the exact moment you can least afford to argue about it, because migrating an HR system with forty people in it is worse than paying.

The honest place to charge

So we asked a different question. Where does our cost actually live? Not in the leave calendar. Not in the org chart, the workflow engine, the documents, or the app on your phone. Those are software: built once, marginal cost near zero, and withholding them from a ten-person company saves us nothing at all. Charging there is charging for the withholding itself.

Our costs live at real boundaries. The payroll module compiles the inputs Capstan already holds and hands your payroll partner a documented export, then files their computed results back onto each record. We run that plumbing; the statutory maths stays with the partner, where it belongs. Contractor work carries country tax forms, compliance documents and currency, all of which have to be kept current in a way a leave calendar does not. Recruiting carries reach and scheduling depth. Infrastructure carries headcount. Enterprise controls such as single sign-on and IP restriction are not built today, and on the day they are, they will carry real support and compliance weight. Each of those costs us something per customer per month, so each costs you something: a module with a published price, or a plan step. Note what is not on that list. The audit trail is not, because every workspace gets it, and charging for the record of what happened would be charging for the withholding again.

That is the whole model. The core HRIS is free up to twenty people, and identically complete on every plan above that. A plan changes two things and only two: how many people you can carry, and whether you can enable modules. It changes nothing about the software, and it does not take the free twenty back: a paid plan carries them, so it charges for the people past twenty. Modules add capabilities that cost us something to run, each with a published price, each cancellable on its own. Nothing in the core is ever moved behind a plan wall to make a quarter.

Free that is not a trap

'Free' in this industry usually means a trial with the clock hidden, or a tier built to hurt until you upgrade, so it is fair to ask what the catch is. The answer is dull. Teams that start at ten become teams of forty, and at person twenty-one you move onto a paid plan at the price on the page. You pay for that person, not for the twenty who came before: the free twenty are carried into the paid plans, so the bill starts at one seat and grows one seat at a time rather than jumping to a whole roster on a single hire. There is no sales call, no quote, and no number that depends on who you are. Until then you are not a lead being warmed, you are a user of a complete product. Some of you will never pay us, and that is fine. The free plan is not generosity. It is the proof, renewed every day we do not gate it, that the model is what we say it is.

The promises on the pricing page are this argument in operational form, and each one is a mechanism rather than a sentiment. The free plan is the full core. A plan gates headcount and module access, never features. Modules are bought and cancelled one at a time. The price you enable at is the price you keep, because the price book is append-only and your entitlement holds the row it was bought against, so a rise reaches you at renewal and never mid-term. A late invoice does not lock you out: access runs uninterrupted for all thirty days, four warnings arrive on the way, and your export keeps working even after the door closes. And your export works on the day you leave, in open formats, whether or not you owe us money.

The same logic runs past the exit. When someone leaves, Alumni Access keeps their own payslips, letters and tax documents self-serve on their personal email, so a document request eighteen months later does not become somebody's afternoon. It is a paid module, billed flat per workspace rather than per person, because alumni are not seats. We would rather charge for it in the open than fold it into a tier and call it free. Most of the category treats a departed employee as a data request to process later. We treat them as a person who should still be able to reach their own records.

What we will not pretend

A manifesto that only lists virtues is marketing. So here is the same honesty pointed back at ourselves.

We are early. Capstan is live in one region, India, and each further region is a separate deployment rather than a switch we flip. Getting onto a paid plan is not self-serve yet: you ask inside the app, and a person here raises the invoice. Single sign-on, SAML and IP restriction do not exist, which is why we are not selling them and why no plan withholds them from you. We hold an ISO 27001 certificate for the way this company works, and we say on the security page exactly what its scope covers and what it does not, because a badge quoted without its scope is a claim wearing someone else's clothes. We hold no SOC 2 report, and we publish no target date for one, because a date nobody has committed to is an intention wearing the clothes of a plan.

Pricing adjusted to purchasing power by country is a policy we have written and not yet finished. Two books are published today, and you can switch between them on the pricing page. That is two out of every country we intend to serve, and the rest are not on the selector, because a book nobody has set is not a price we would honour.

None of that is here as a confession. It is here because a company that will tell you what it has not built yet is the same company you can believe about what it has.

What we refuse

A pricing position is also a product position, and we refuse dark patterns in both. No 'contact sales' wall around a number we already know. No exit-intent popups, countdown timers or invented scarcity. No gated PDFs. No chat widget pretending to be a person. And no AI passing judgment on your people: nothing here scores a person, ranks them against their colleagues, or decides something about them without a rule you can read. What the software determines, an entitlement, an access decision, a line in a letter, comes from that rule rather than a model's guess, and your people's data is never sent to an outside model or used to train one. Software that evaluates humans in the dark is not a feature, whatever the category's marketing says this year.

None of this is charity, and we would rather you did not read it as idealism. It is a bet: that the money spent engineering artificial scarcity, the packaging meetings, the gate maintenance, the sales choreography, is better spent making the product good and the price legible, and that founders talk to each other. If we are right, transparent pricing is not a margin sacrifice. It is the cheapest marketing there is.

So the pricing page is public and it is complete. Every number we charge is on it, and when a number moves, the change is announced before it takes effect and dated in public, so the page accumulates a history you can check rather than a quote you have to request. That sentence is the product strategy, the marketing strategy and the promise, all at once. Hold us to it.

The Capstan team, PeopleCap
August 2026