Every agency has sat through a discovery call that should have been a five-minute email exchange instead: the prospect wants a full brand redesign on a $2,000 budget, or a "quick landing page" that turns out to need six integrations, or a launch date that was already unrealistic before the call started. None of that needed to be discovered live. It needed to be caught by the form before the calendar invite went out.
A "book a discovery call" button doesn't catch any of it. It's a scheduling tool, not a filter — it books everyone who's curious with the same weight as everyone who's ready to sign. The fix is the same one that works for any service business: put a real qualification step in front of the booking link, one that branches on what the prospect actually needs and scores whether their budget matches their scope before a human ever opens a calendar.
Why "tell us about your project" doesn't work
Most agency intake forms have one open-ended field — "tell us about your project" — and call it qualification. It isn't. An open text field gets either a two-word answer ("need a website") or three paragraphs that still don't answer the two questions that actually predict whether this is a good-fit client: what type of project is this, and does the stated budget match what that type of project actually costs.
Those two questions can't be answered by one field. They need branching: the project type a prospect selects should change what budget ranges are even shown next, and the combination of the two — not either one alone — is what determines fit.
A real intake flow, question by question
Here's a flow built for a design/dev agency as a concrete example — the structure holds for any project-based agency, adjust the specifics to your services:
- "What kind of project is this?" — Multiple choice: new website, brand identity, ongoing design retainer, one-off campaign, other. This is the branch point everything else depends on — a website build and a monthly retainer have completely different budget realities and should never share the same next question.
- A branching budget question specific to their first answer. Someone who selected "new website" sees realistic ranges for a website ($5–15k, $15–40k, $40k+). Someone who selected "brand identity" sees ranges scaled to that instead. Showing the same generic "what's your budget?" range to everyone is exactly what produces the mismatched-expectations call.
- "When do you need this live?" — "Within 4 weeks," "1–3 months," "No fixed date yet." Combined with project type, an unrealistic timeline for the selected project type is itself a strong signal, independent of budget.
- "Have you worked with an agency before?" — First-time buyers and repeat agency clients need different framing in the outcome screen, even when their budget and scope both check out.
- The outcome, scored, not generic. A good budget-to-scope match sees a booking link for a real discovery call. A promising but underfunded prospect sees a smaller starting package instead of the full engagement they asked for. A clear mismatch sees a resource — a project cost guide, a DIY starting point — instead of a dead end.
What actually predicts a good-fit agency client
The signal that matters most for agency work isn't budget alone — it's whether the stated budget is realistic for the stated project type. Three things to weight:
- Budget-to-scope match, not budget in isolation. A $10k budget is a great fit for a landing page and a poor fit for a full e-commerce rebuild. Score the combination, not either field alone — this is exactly the kind of condition a flat form can't express, but a branching rule can: "if project = 'website' AND budget < $5k, route to the DIY resource," while the same budget under "brand identity" might be a fine fit.
- Timeline realism against project type. "4 weeks" for a brand identity project is a mismatch before anything else is even discussed; "4 weeks" for a one-off campaign might be completely normal.
- Decision-making authority. Whether the person filling out the form is the actual budget-holder or is "gathering quotes for the team" changes how fast this can realistically move — worth a direct question, since it changes the outcome screen's next step (book a call vs. send a proposal to share internally).
Keep the question count low — the branching does the qualifying
Five core questions is enough when two or three of them branch based on the previous answer. The temptation is always to ask more questions to qualify harder; completion rates drop fast past 6–8 questions, and an abandoned form filters nobody. A shorter form where the budget question actually adapts to the project type produces sharper signal than a longer form that asks every prospect the same generic range regardless of what they're building.
The form doesn't need to be long to qualify well — it needs the second question to change based on the first.
Building this without a developer
Project-type branching, a scored budget-to-scope match, and a personalized outcome screen per fit tier are all buildable visually in an afternoon — no code, no separate scoring spreadsheet after the fact. The free plan covers this exact use case with no cap on branching complexity, and the outcome screen can hand a non-fit prospect a genuinely useful resource (a project cost guide as a branded PDF) instead of a plain "we'll be in touch," which is what actually protects the relationship with the leads you turn away.