What Does a Discovery Phase Before a Development Project Cost?
A discovery phase before a development project normally costs SEK 100,000–300,000 and takes 3–6 weeks. The price depends on the project's size, the number of stakeholders, and how much technology needs investigating. The result is a target vision, solution proposal, risk map, and an estimated backlog – input that makes the rest of the project cheaper and more predictable.
A discovery phase is the cheapest insurance against the most expensive thing that can happen in a development project: building the wrong thing. Yet it’s often the line item clients try to cut first. Here’s what a discovery phase costs in 2026, what you get for the money, and how it pays for itself.
The price picture: SEK 100,000–300,000 over 3–6 weeks
A discovery phase before a development project normally costs SEK 100,000–300,000 and runs 3–6 weeks. Where you land in the range depends on how much needs investigating.
| Scope | Typical cost | Timeline |
|---|---|---|
| Contained product with a clear idea | SEK 100,000–150,000 | About 3 weeks |
| Multiple stakeholders and integrations | SEK 150,000–250,000 | 4–5 weeks |
| Complex environment with many systems and strict requirements | SEK 250,000–300,000 | 5–6 weeks |
What drives the cost is the number of stakeholders to interview, the number of existing systems to map, and how large the technical uncertainty is. A new standalone service is cheaper to investigate than a product that has to connect to three internal systems.
The deliverables: what you should get
A discovery phase that only produces a PowerPoint full of visions isn’t worth the money. Demand concrete deliverables:
- Target vision – what the product should achieve, for whom, and how success is measured.
- Solution proposal – overall architecture, technology choices, and how the solution fits into your existing environment.
- Risk map – the biggest uncertainties ranked, with proposals for how each is handled or tested.
- Estimated backlog – the product broken down into building blocks with rough estimates, prioritized by value.
With these four pieces in place, you can make an investment decision, gather comparable quotes, and start development without a long ramp-up.
How the discovery phase lowers total cost
The discovery phase’s economic value lies in two things. The first is the right scope: once the target vision is clear, you can cut features that don’t contribute to it. It’s common for the estimated backlog, after prioritization, to end up noticeably smaller than the original wish list – features that are never built are the cheapest features there are.
The second is fewer do-overs. A do-over is when something gets built, turns out to be wrong, and has to be redone. That’s the most expensive way to work in a project, since you pay for the same thing twice and lose calendar time on top of it.
Worked example: what does a do-over cost?
Say a team of three builds a module over four weeks, and only afterward does it turn out to solve the wrong problem. Three people for four weeks is about 480 hours. At an agency hourly rate of SEK 900–1,400, the do-over has cost SEK 430,000–670,000 – plus a month’s delay.
A discovery phase costing SEK 150,000 therefore only needs to prevent a fraction of one such do-over to pay for itself. In practice it often prevents several, since most do-overs come from unclear goals or unknown technical obstacles – exactly what the discovery phase exists to flush out. Think of it as part of the development work, not a cost on the side.
Three requirements to set for the vendor
- Fixed price. A discovery phase is well contained and should be quotable at a fixed price – it’s also an excellent test of the vendor’s ability to estimate.
- Vendor-neutral deliverables. The material should be yours, in a format you can pass on, and written so another vendor can quote against it.
- A recommendation with backbone. Require a clear stance: build, trim the scope, or walk away. A discovery phase that buries the conclusion in appendices hasn’t done its job.
When can you skip the discovery phase?
A discovery phase isn’t always justified. If the project is small, the domain well known, and the team experienced, one or two workshops are often enough to create alignment. The rule of thumb: when the investment is under half a million and the risks are known, you can go straight to the build. When the investment is in the millions, several systems are involved, or stakeholders don’t agree – then SEK 100,000–300,000 for a discovery phase is cheap risk management.
At Weapp we start most larger engagements with some form of discovery phase, precisely because the projects that follow move faster and stay on budget. Want to know what a discovery phase would cover for your idea? Get in touch and we’ll give you a concrete plan with a price.
Frequently asked questions
What's the difference between a discovery phase and a design sprint?
A design sprint is a compressed week that tests a specific concept with users. A discovery phase is broader: it investigates goals, technology, risks, and cost over 3–6 weeks and produces a complete decision basis. Many projects use both – the discovery phase for the big picture, a sprint for the most uncertain concept.
Can I use the discovery phase to get quotes from several vendors?
Yes, a good discovery phase is written vendor-neutral. The target vision, solution proposal, and estimated backlog serve as a request-for-quote package, and the quotes become comparable because everyone is pricing the same thing. Require that the material be delivered in a format you own and can pass on.
Does the same agency that builds the product have to run the discovery phase?
No. The advantage of using the same vendor is that the knowledge carries straight into the build without a handover. The advantage of separating them is an independent assessment of what should be built. Both approaches work – what matters is that the deliverables are concrete enough to build on.
What happens if the discovery phase shows the project shouldn't go ahead?
That's a successful outcome, even though it rarely feels that way. You've paid SEK 100,000–300,000 to avoid spending millions on the wrong venture. A discovery phase that just confirms what everyone already thought has done a worse job than one that changes the decision.
How long does a discovery phase stay valid before it goes stale?
As a rule of thumb, 6–12 months. The market, the organization, and the technology landscape have time to change, and the estimates rest on assumptions that need revisiting. If the discovery phase is older than that, it should be updated before it's used as a decision basis – a review is much cheaper than a new discovery phase, though.