Why a fixed quote before discovery is a red flag

Every few weeks someone asks for a fixed number before we've talked about what they're actually building. It's a reasonable thing to want — budgets need bounds. But a number given before discovery is a guess wearing a suit.

Software estimates are only as good as the assumptions underneath them, and most of the expensive assumptions don't surface until you've actually mapped the data model, the integrations, and the edge cases a client didn't think to mention because they didn't know they mattered. A studio that skips that step and hands you a number anyway is either padding heavily to cover the unknowns, or planning to renegotiate the moment reality intrudes.

Discovery isn't a sales formality before the "real" scoping call. It's the only point in the process where the risk is still cheap to find. A week of discovery that surfaces a hard integration problem is worth more than a month of development that discovers it the hard way.

What happens during discovery?

A focused discovery covers:

• The outcome: what has to be different for the business when the project is done, and how it will be measured. See why we scope outcomes, not features.
• The users: who will use it, what they do today and where they get stuck.
• The data: what records exist, where they live and how they relate.
• The integrations: every system the project must talk to, and how good their APIs actually are.
• The edge cases: refunds, cancellations, duplicates, permissions, the things that break real products.
• The constraints: deadlines, budget, compliance and who on your side makes decisions.

What should you get at the end of discovery?

• A written outcome statement.
• A scope listing what is in and what is out.
• The main risks and how they will be handled.
• A timeline with milestones you can see working.
• A fixed price for that scope, and how changes will be quoted.

If discovery ends with only a verbal estimate, it was a sales call.

How long should discovery take?

For a small business website, a short call and a checklist are often enough. For a custom web app or an AI integration, discovery usually takes days to a couple of weeks, depending on how many systems and stakeholders are involved. It should be proportionate: long enough to find the expensive unknowns, short enough that you are not paying to plan forever.

Is a quote ever reasonable before discovery?

A range can be, as long as it is labelled as a range and the assumptions are written next to it. Published starting prices also help you decide whether a conversation is worth having. Ours are on each service page, such as website redesign and custom web app development. What is not reasonable is a firm number for a complex project from someone who has not asked the questions above.

How do you compare quotes from different studios?

Put them side by side against the same scope, and check what each one includes: discovery, design, testing, launch support, documentation and ownership of the code. The cheapest quote often leaves one of those out. We go further in how much does custom software development actually cost in 2026? and how to choose a software development partner.