How we price a project when the client doesn't know the scope yet

A prospective client asked us for a fixed number on a call recently, for a project she herself admitted she couldn't fully describe yet: "something like a customer portal, but I'm not totally sure what it needs to do." We could have guessed a number to keep the conversation moving. We told her instead that anyone giving her a confident fixed price at this stage was guessing too, whether they said so or not.

This happens more than people might expect. Plenty of real, fundable projects start life as a rough idea rather than a fully specified brief, and pretending otherwise to produce a tidy number doesn't serve either side. What we do instead is price the part we can actually see clearly, and treat the rest as a separate, smaller decision.

When scope is genuinely unclear, we quote a short, fixed-price discovery phase on its own: usually one to two weeks, aimed specifically at producing a real scope document, a rough architecture, and a defensible cost estimate for the full build. It's a small, contained commitment for the client, and it's honest about what we don't yet know.

This isn't a sales trick to get a foot in the door, though we understand why it can look that way from outside. The discovery phase has to stand on its own value, producing something the client could hand to a different agency entirely if they chose to. We tell clients this explicitly, because the offer only means something if it's genuinely true.

A written scope document specific enough that any competent agency could price against it without needing to ask the founder clarifying questions first. A rough technical architecture identifying the hard problems early, before they turn into expensive surprises mid-build. And a cost range for the full project, with the assumptions behind that range written down explicitly rather than left implicit.

We've had clients take this document to another agency for a competing quote, and we're fine with that outcome. If the document is genuinely good, it's useful regardless of who ends up building the actual project, and we'd rather earn the follow-on work by being the obvious best fit than by making it awkward to leave.

A number given without real scope tends to be either too high, scaring off a client who might have been a great fit at an accurate price, or too low, setting an expectation that gets painfully renegotiated later once real scope is known. Neither outcome serves the relationship, and the renegotiation conversation in particular tends to damage trust that's hard to rebuild afterward.

We've watched competitors win early-stage deals with confident low numbers that later ballooned once real scope emerged, and we've watched the resulting client anger land squarely on the agency, not on the vagueness of the original brief. Being the agency willing to say "we don't know yet" out loud has cost us a few deals up front and saved us from several bad ones later.

It's a harder pitch than a confident number on the first call, and it requires a client willing to invest a little time and a little money before getting the full picture. Some prospects walk away at this point, usually the ones most focused on getting a cheap number quickly rather than an accurate one they can actually plan around.

The clients who stay tend to be better clients overall, because agreeing to a discovery phase is itself a small signal of how they'll approach the rest of the relationship: willing to invest in getting things right rather than rushing toward a number that feels good in the moment and turns out to be wrong later.

Sometimes the discovery phase surfaces a project meaningfully larger than either side expected going in. When that happens, we present the real number plainly, with the reasoning laid out clearly, and let the client decide whether to proceed with the full scope, phase it into smaller pieces, or walk away with a document that's still useful to them regardless.

We've had clients choose all three of those paths, and we consider each one a legitimate outcome of an honest discovery process. The alternative, a fixed number given too early that turns out to be badly wrong, doesn't actually avoid this conversation. It just delays it to a worse moment, after real money and real time have already been spent.

Sometimes a client needs a number for an internal budget approval process and can't wait for a formal discovery phase to complete. In that case we give a wide range, explicitly labeled as a range built on assumptions, with the specific assumptions listed next to it, so nobody mistakes it for a firm commitment later.

We've found that labeling the uncertainty honestly, rather than either refusing to give a number at all or quietly rounding to something falsely precise, tends to satisfy the internal approval need while protecting both sides from a number that was never meant to be treated as final.

"We don't know yet" is an uncomfortable thing to say in a sales conversation, and it costs us some deals with clients who wanted certainty more than they wanted accuracy. We've made peace with that trade.

The clients who do move forward on this basis tend to be the ones we do our best work for, because the relationship starts on a foundation of honesty about what's actually known, rather than a number neither side can really trust once the real work begins.

A few of them have told us afterward that the discovery phase itself was the moment they decided to trust us with the bigger build, not the pitch that came before it. That's usually a better signal of fit than any proposal deck could have given either of us at the start.

We've started offering this same structure by default now, even to clients who arrive with a scope they're confident is complete, because a short review pass has caught real gaps often enough to be worth the extra day it adds before a full number goes on paper.