How Much Does a Discovery Phase Cost in Software?

October 9, 2026 · 12 min read

A software discovery phase usually costs somewhere between a few thousand dollars and several tens of thousands, and large regulated projects go well beyond that. The honest answer is a calculation, not a number: the people on the team, times the weeks, times their weekly rate. This post shows the published figures, why they disagree, and how to work out a range you can defend, whether you are buying a discovery or pricing one.

For the steps and deliverables, see our guide to the discovery phase in software development; this post goes deeper on the money.

Typical price ranges and why they disagree

Here is what a few agencies publish about their own discovery work. These are the vendors' own figures, for their own scope and rates, not market surveys.

Read side by side, an MVP discovery costs anything from $2,500 to $40,000. That is not because some vendors are wrong. They are pricing different things:

• Different work under the same name. One firm's discovery is two weeks of workshops and a backlog. Another's includes user interviews, architecture, wireframes, a roadmap and a fixed-price build offer. Same label, several times the effort.
• Different rates. The same ten person-weeks cost very different amounts depending on where the team sits and how senior it is.
• Different team sizes. A single analyst for two weeks and a team of four for two weeks are both "a two-week discovery".
• Different commercial intent. A low fixed fee, or a fee deducted from the build, is partly a sales tool: the vendor expects to recover the cost in the build that follows. A high fee may include work other firms do in the first sprints.

The percentage rule of thumb that agency guides repeat sits on top of all this. easy.bi, for example, puts discovery at 5-10% of the total project budget, and our pillar guide uses a wider 5-15%. Treat the percentage as a sanity check, not a price: on a large build the share falls, on a small risky one it can be higher.

What drives the cost

Underneath every quote is effort, and effort comes from a short list of factors. When two proposals differ a lot, check them against this list first.

• Scope of the question. "Can we replace this spreadsheet process?" is a smaller discovery than "Should we build a platform for three markets?".
• Number of user groups. Each distinct group (customers, field staff, back office, admins) needs its own interviews, journeys and requirements. This is often the biggest single driver.
• Integrations. Every system the product must talk to has to be checked: its API, its data, who owns it. Poorly documented or legacy systems take far longer than a modern, well-documented API.
• Compliance and data. Health, finance, personal data and data residency add reviews, non-functional requirements and sometimes a specialist on the team.
• Team composition. A business analyst alone is cheaper than a business analyst, designer and architect, but the cheaper team cannot check integrations or sketch screens. Know what you are giving up.
• Decision speed. Calendar time is usually lost waiting for interviews and answers. If the client needs a week to approve each decision, a fixed-fee vendor will price that delay in, and a time-and-materials one will bill for it.
• Depth of design. Low-fidelity wireframes are part of most discoveries. Polished visual design inside discovery adds effort without answering the risky questions.

A worked cost model

The most useful way to price a discovery is in person-weeks:

cost = people on the team x weeks x blended weekly rate

The effort figures below are the same rules of thumb as in our pillar guide. The rates are illustrative assumptions chosen to show the spread, not survey data: replace them with your own or your vendor's.

For orientation, $1,500 a week is $37.50 an hour at 40 hours, and $4,000 a week is $100 an hour. Many agencies charge more than that, so the top column is not a ceiling.

You can also run the model backwards to read a quote. Divide the price by a plausible blended rate and you get the effort the vendor has priced in. If a four-week discovery comes to two or three person-weeks, either the team is one person part-time, the scope is narrow, or the fee is partly recovered in the build. Know which one you are buying.

Two adjustments worth adding to your own model:

1. Client time. Your sponsor, product owner, users and IT staff will spend hours in interviews and reviews. It does not appear on the invoice, but it is part of the cost and the most common cause of delay.
2. A contingency. If the scope is unclear going in, allow 10-20% on top, or agree a cap (see below). This is a rule of thumb, not a measured figure.

Pricing models: fixed fee, capped time and materials, credited, or "free"

How a discovery is priced matters as much as the number.

Whatever the model, the contract should say what you receive, when the phase ends, and that the documents are yours to take to another team.

What you should get for the money

A discovery is worth its price only if it ends with documents you can act on. The pillar guide lists the ten deliverables a finished discovery should leave: stakeholders, background, executive summary, problem statement, goals and success metrics, user experience, requirements, architecture, UI screens, a roadmap with ranged estimates, plus a risk log.

The requirements are usually the part that most affects the build price, so check them hardest: each one testable, prioritized, with acceptance criteria and a link to a goal. Our software requirements document template shows that structure filled in.

To see what all of it looks like together, the example discovery report shows a complete set for Fieldwise, an invented field-service app; the names and figures are made up, the structure is the point.

How to keep the cost down without skipping the work

Most of the savings come from removing waiting and rework, not from cutting stages.

• Send everything you have before kickoff. Decks, old specs, process documents, support tickets, interview notes. The team should not spend paid days rediscovering what is already written.
• Book interviews before the phase starts. A discovery that waits a week for the first user interview has burned a quarter of a four-week budget.
• Name one decision maker. Someone with real authority who can answer within a day or two.
• Narrow the question. Discover the first release in depth and leave later phases rough.
• Keep design low-fidelity. Wireframes answer the questions; visual polish can wait.
• Use a standard structure. Working from a list such as the discovery phase checklist saves the team from inventing a format and makes gaps visible early.
• Draft from existing material. Turning decks, specs and call transcripts into first drafts is slow, repetitive work; draft it once and spend the paid hours on review and decisions (see the next section).

If you run discovery: pricing your own

If you are the agency, consultant or in-house team doing the discovery, the same model works from the other side. A fixed fee is just your person-week estimate, times your rate, plus a contingency for the unknowns you cannot remove before the start.

The useful question is where those person-weeks go. Roughly, the work splits into two kinds:

How large the second share is differs by team and project, so check it against your own timesheets. It is also where rework piles up: one late change to a goal or requirement has to be carried through the architecture, the screens, the estimate and the risk log by hand. When you price a fixed fee, look at your own past projects and estimate how many of the person-weeks went into writing and assembling rather than talking and deciding. That is the part you can shorten without cutting the discovery itself.

Three practical consequences for your proposal:

1. Price the deliverables, not the meetings. List the documents the client will receive and what "done" means for each; it makes your fee comparable and defensible against a cheaper quote that delivers less.
2. Show a finished example. A sample report, with the client's name nowhere in it, answers "what do we get for this money?" faster than a description. The example discovery report shows what a complete set looks like.
3. Keep the drafting time out of the critical path. If first drafts of requirements and journeys exist the day after the interviews, the review meeting can happen that week instead of the next.

This is the part Discovery Phase AI was built for: it drafts the early stages (background, problem, stakeholders, goals, users and journeys, requirements) from the decks, documents and transcripts you already have, helps you write and score the rest, and exports the result as one report. Your team still runs the interviews, reviews every draft and makes the decisions; what changes is how many hours go into writing it all down.

Questions to ask before you sign

Whether you are the buyer or the agency writing the proposal, these questions expose most of the difference between two quotes:

1. Which named documents will we receive, and what does "done" mean for each?
2. Who is on the team, for how many days each, and at what rate?
3. How many user interviews are included, and with which groups?
4. Which integrations will be checked, and by whom?
5. Is the estimate at the end a range with written assumptions, or a single number?
6. What happens if the phase runs over: change request, cap, or extra invoice?
7. Is the fee credited against the build, and do we own the deliverables if we go elsewhere?
8. What do you need from us, and by when, for the timeline to hold?

FAQ

How much does a discovery phase cost for an MVP?

Published vendor figures for an MVP range from about $2,500 to $40,000. With the model above, a two-week discovery of 3-4 person-weeks comes to roughly $4,500-$16,000 at the illustrative rates; your own rates decide where you land.

Is a discovery phase worth paying for?

When the build is large or uncertain, usually yes: a few person-weeks of discovery cost far less than building the wrong thing. When the change is small and well understood, keep it to a day or two of writing down the problem, goals, Musts, integrations and top risks.

Why is one agency's discovery five times cheaper than another's?

Usually because it covers less work, uses a smaller or lower-rate team, or is partly paid for inside the build. Convert both quotes into person-weeks and deliverables before comparing.

Should discovery be a percentage of the build budget?

Agency guides often quote 5-10% or 5-15%. Use it as a check on your person-week estimate, not as the price itself.