The budget conversation: why hiding the number costs you more

There is a conversation that happens in almost every first call with a prospective client, and it goes like this. We ask what budget they're working with. There's a pause. Then some version of: "We'd rather see what you propose first."

We understand the instinct. If you've ever watched a tradesperson's quote rise to meet the number you mentioned, the logic feels airtight: whoever names a number first loses. But agency work isn't a commodity with a sticker price — it's a design problem with a budget as one of its inputs. Hiding that input doesn't keep the price honest. It just guarantees the agency designs for a fictional budget, which produces one of two outcomes: a proposal for the project you can't afford, or a proposal padded against the ambiguity you created. Either way, you've bought yourself two wasted weeks and a worse price than candour would have got you.

Here's the case for naming the number, from the side of the table that receives it.

What an agency actually does with your number

When you tell us the budget is — say — $120k, three legitimate things happen, and none of them is "great, that's the invoice."

First, we design the right-sized response. A $120k answer and a $300k answer to the same brief are different projects, not the same project at different margins. At $120k we might propose a fourteen-week sprint: senior squad, focused scope, opinionated defaults, your content team owning the copy with our editorial support. At $300k the same ambition gets a discovery phase, a research program, motion design, a custom imagery language. Both are good projects. The mismatch is what's expensive — a $300k-shaped proposal in front of a $120k client is dead on arrival, and we've both spent real hours on a corpse.

Second, we tell you when the number is wrong. This is the part clients don't expect from a sales conversation: a decent agency will regularly say "you don't need to spend all of that" or, more often, "that scope won't fit that number — here's what version does." The projects we are proudest of include several where we talked the client down — a smaller first phase with a measurable goal, then a second phase funded by evidence. An agency that never says "spend less" is telling you something about their model.

Third, we can route you honestly. If the budget is $40k and the brief needs $150k, the respectful answer is immediate: here are the three things you can do in that range, here are two good smaller studios for the rest, come back for phase two. Every week of dance concealing that gap is a week off your launch date.

Why hiding the number makes quotes worse, not better

Procurement theory says concealed budgets create competitive tension and compress prices. In practice, for bespoke work, the compression shows up as risk premium, not savings. When the budget is unknown, the quoting team must defend against the worst case: the client who wants the moon and will punish every omission in the final invoice. So the estimate grows contingency, assumptions get conservative, and scope gets padded with items you're unlikely to need — all of which you pay for.

Transparency does the opposite. A named number lets the agency spend its margin openly: "Here's what fits exactly, here's the trade at the edge, here's the contingency we recommend and why." You can audit a quote built this way line by line. You cannot audit the fog a mystery budget produces. (Our piece on how studios price and how to read a quote decodes the line items in detail — worth reading before your next proposal round regardless of who you hire.)

There's a second-order cost too: schedule. Every serious proposal takes a week or two of senior time to write properly. Churning through proposal rounds because each one whiffs the hidden budget is the slowest possible way to start a project, and starting late is the most reliable budget overrun there is.

What the ranges actually buy

Numbers below are illustrative — rates vary by market and studio — but the shape of what money buys is stable, and understanding it makes the budget conversation easy.

• Under $50k: you are buying a sharp, bounded thing. One landing ecosystem, a brand refresh with tight deliverables, a design-system codebase on top of an existing product, a focused CRO program. Great money if genuinely focused; a trap if the brief secretly contains three projects.
• $80k–$200k: the workhorse range. A complete marketing site with a real CMS and a coherent design system; a product redesign of one core experience; a headless commerce build with sane integrations. This is where the quality-per-dollar is usually highest — enough money for senior people and a real process, not enough for process theatre.
• $250k–$600k: a platform. Brand plus site plus design system; a SaaS product from validated concept to launch; a commerce replatform with subscription logic, internationalisation and a content operation behind it. At this range you're buying a team for a quarter or two, and the engagement model matters as much as the number — see retainer or project?.
• Beyond that: you're buying a partnership, and you should evaluate the relationship as hard as the capability, because you'll be in it for years.

The pattern worth noting: within every range, money buys range of problem, not pixel quality. The $80k site's typography can be as good as the $400k one. What the extra money buys is the number of problems that can be excellent simultaneously — research depth, motion, content, performance engineering, the tenth template instead of the fourth.

If you genuinely can't name a number

Sometimes the silence is real, not tactical: the budget isn't approved yet, or it spans departments and nobody owns the sum. In that case, name a band: "somewhere between $80k and $150k, and we'd phase it if it trends high." A two-times band is close enough to design against, and honest about its own uncertainty.

If procurement rules genuinely forbid disclosure, say that, too — "we can't share the number, but we can tell you we've rejected two proposals as unaffordable and one as suspiciously cheap" is real signal. Silence is the only useless answer. Even a refusal is information if it's explained: a good agency will then propose tiered options (a lean, a right, a maximal version of the same scope) — which is slower to write, and, yes, priced into the proposal, which is why directness is cheaper.

Negotiate scope, not rate

This is the negotiation advice we give friends who hire our competitors: never beat an agency down on rate. Not for ethical reasons — for mechanical ones. A studio that agrees to your budget against its rate card has only four places to find the difference, and you won't like any of them: more junior staff than the proposal implied, fewer review rounds than the work needs, shortcuts in the invisible layers (accessibility, testing, performance budgets), or a resentful team doing exactly what was specified and nothing the work deserves. The deliverable list doesn't change; everything inside the deliverables quietly does. You cannot write a contract clause that forces care.

Scope, by contrast, negotiates cleanly. "The budget is firm at $140k" produces a useful conversation: drop the blog migration (do it in-house with our templates), defer the member dashboard to phase two, use a system typeface this year and commission the custom one next. You're trading what gets built, and every trade is visible and choosable. The same dollar conversation, conducted as rate-haggling, trades invisible quality you won't discover until launch week.

Two scope moves deliver outsized value for the money they save. Cutting breadth before depth — nine great templates beat fourteen adequate ones, every time, in both outcomes and maintenance. Buying discovery separately — a fixed-scope discovery sprint before the big commitment shrinks the fat end of the estimate, because most contingency is fear of the unknown and a sprint converts unknowns into a priced plan.

The version of the conversation that works

The whole thing, done well, takes four minutes of the first call: "Our budget is $X, approved, with $Y held for contingencies we control. If the honest version of this brief doesn't fit, we want to hear that in the first round, with options — not in change orders in month three. What fits?"

Every credible agency will exhale, tell you the truth, and spend the next proposal on your project instead of on your mystery. The ones who respond to that candour by inflating to the number have just told you, cheaply and early, not to hire them. Which is, in the end, what the budget conversation is really for: not setting a price, but running the first honesty test of the relationship — and you'd rather run it before the contract than after. When you're ready to have it with us, the project brief form takes about ten minutes, and our engagement models are written in the same dollars-first spirit.

Key takeaways

• Budget isn't a negotiation chip to conceal — it's a design input. Hiding it buys a proposal for the wrong project, padded against the ambiguity you created.
• A named number enables the three behaviours you actually want from an agency: a right-sized proposal, honest "spend less / this won't fit" advice, and fast routing when the gap is real.
• Unknown budgets produce risk premium, not savings. Transparency is what makes a quote auditable line by line.
• At every range, money buys breadth of problem, not pixel quality. Know the shape of the ranges so you can size your brief realistically.
• Negotiate scope, never rate. Rate-haggling extracts invisible quality you can't contract back; scope trades are visible and choosable.
• Discovery-first is the highest-value cut: it converts contingency (which is fear of unknowns) into a priced, evidence-backed plan.

FAQ

Won't the agency just quote to the top of whatever number I give? Good ones don't, because it's a bad trade: an inflated quote loses competitive rounds and torching trust in round one kills the expansion work agencies actually live on. But here's the beauty of disclosure — it makes inflation detectable. A quote built to a named budget shows its working; ask "what would the $90k version sacrifice?" and a padded quote has no coherent answer, while an honest one lists real trade-offs immediately.

We got three wildly different quotes for the same brief. Is someone lying? Usually nobody is — they're answering different questions. Read the assumptions section of each before comparing totals; the cheapest quote typically assumes your team writes all content, supplies all imagery, and accepts the platform-as-is. Normalise scope first, then compare. The quote-reading playbook walks the diff line by line.

Is it ever smart to genuinely run a blind competitive proposal? For commodity work — hosting, licences, printing — yes. For bespoke creative and engineering work, blind rounds select for the best guessers, not the best builders. If procurement mandates them, publish a band and score proposals on how well the team reasons about the trade-offs inside it; that single change improves the quality of every response you'll receive.

What if our number is embarrassingly small — do we say it anyway? Especially then. Small numbers get an honest answer fastest: either "here's the sharp, bounded thing that budget does brilliantly" or "here's who does this well at that scale." The embarrassment is imaginary; studios see the full spread of budgets and respect the client who knows theirs far more than the one performing scale they don't have.