A paid discovery phase is worth it when it ends with something you own and could hand to a different supplier. It is not worth it when all you get is a proposal from the company that ran it.
That distinction matters more than the fee. Discovery, sometimes called scoping, is sold by almost everyone and means almost nothing on its own.
Key takeaways
- Discovery is where a reliable price comes from. A fixed price quoted without it is a guess with a confident face on it.
- For most small and medium projects, proportionate discovery is one to two weeks of senior time, which puts it in the low thousands of pounds rather than the tens of thousands.
- A few hours of free pre-sales work is normal and can be genuinely useful. It is not a discovery phase, and the difference is measured in weeks of work, not in whether anyone invoiced you.
- A supplier being paid for discovery can afford to tell you not to build the thing. One doing it free, hoping to win the build, cannot.
- Discovery is not worth paying for when the job is small enough to describe on a page, or when it would swallow a serious share of your build budget.
What you should be holding at the end
The test is portability. If you could hand what you have to three other suppliers and get comparable quotes back, discovery did its job. If it only makes sense as a preamble to hiring the people who wrote it, you have bought a sales pitch and paid for the privilege. Two honest shapes pass that test.
If you want to know what a build will cost, you should end up with a written specification saying what the software does, who uses it and what is out of scope, some designs or wireframes, because most people cannot react usefully to a description but can react immediately to a picture, and a price against that specification. Ask how long the price stands. A quote with no shelf life is not much of a commitment, and if it expires in a fortnight you cannot really take it anywhere.
That is the shape behind every build we can quote firmly. On Benefits Buddy, a staff-benefits platform with several types of user across web and mobile, the scope was written down in phases before any code was written. The price was firm because the specification came first, not the other way round.
If you intend to go out to market, you should end up with a findings report and a requirements document you can put out to tender. The findings say what you are running now, what is wrong with it, and what your realistic options are, covering your processes and not only your software, because a fair share of what feels like a technology problem is a working-practice problem no new system will fix. The tender document turns that into functional, technical and implementation requirements any supplier can bid against. That is the most portable form discovery takes, because it is built to be sent to your supplier's competitors. Work of that shape is closer to independent technology consultancy than to the front end of a build, and it is worth buying it that way, from someone whose only job is the recommendation.
It is also why the price cannot come first. Nobody can commit to a firm number for work nobody has pinned down, which is why fixed-price projects start with a specification rather than a quote.
What it should cost, and what drives it
Most of what you will find online is priced in dollars, or quoted by firms who will not say what the number covers. So, plainly: for a typical small or medium business project, a proportionate discovery is a couple of weeks of work by a small senior team, and lands in the low thousands of pounds rather than the tens of thousands. For context on the rates behind that, see what bespoke software costs in the UK.
What moves it is not the size of the eventual build. It is how much there is to find out, and how many people you have to find it out from. Interviewing half a dozen staff across a couple of days on site is a known quantity. Untangling how six sites currently do the same job in six different ways is not.
A rule of thumb we would apply as a buyer: if discovery is being quoted at more than roughly a tenth of what you expect the build to cost, ask what makes it that big. There may be a good answer. There may not.
Is free discovery worth having?
Yes, as long as you know what it is. Plenty of suppliers give work away before you have committed to anything, and we do a version of it ourselves: a short document, sometimes a few wireframes, a few hours of thinking. That is a normal part of selling and it is worth having as a buyer. You will learn more about how a supplier thinks from a few hours of their work than from any number of case studies.
What it is not is a discovery phase. Nobody has interviewed your staff, looked at the systems you actually run, or written anything you could put in front of a different supplier. The distance is measured in weeks of work, not in whether an invoice was raised, and a supplier who blurs the two is hoping you will not notice until later.
Free work is also paid for somewhere: recovered in the build price, or paid in the quiet obligation you feel towards people who have just given you their time for nothing. Neither is sinister. Both are worth seeing.
The structural point is about incentives. A supplier being paid for the discovery itself can afford to reach an inconvenient conclusion: that an off-the-shelf tool would do, that some of the problem is process rather than software, that the project should wait. A supplier doing it free, hoping to win the build, has an obvious reason not to.
We have been paid to scope projects where our recommendations began with getting more out of the systems the client already had, and changing how some things were done, before any new software was mentioned at all. That is a much easier report to write when writing it is the job you are being paid for.
When discovery is not worth paying for
If the job is small and you could describe it on a page, do not buy a process to tell you what you already know. Ask for a quote and get on with it. The same goes if you have already been through this with somebody else and own a specification: take what you have and get it priced. We will quote a build against a specification we did not write, and so should anyone else worth hiring. Being told you have to buy their discovery first, when you are holding a perfectly good document, is a sales position rather than a technical one.
If discovery would eat a serious share of your total budget, the answer is not to skip understanding the problem, it is to buy less of it. A couple of days that produce a rough shape and an honest range is better than a fortnight you cannot afford. Any supplier worth hiring will scale it to what is in front of them, and one who will only sell the full engagement has told you something useful.
What to do next
Before you agree to a discovery phase, ask one question: what will I be holding at the end that I could take to someone else? A supplier who answers that easily is selling you an asset. One who talks around it is selling you a step in their sales process.
Our own scoping service is built to pass that test. We interview your key staff, then hand you a detailed specification, interactive wireframes and a fixed-price quote that stands for six months. All of it is yours, and you are free to take it to any supplier.
If you are not sure your project needs one at all, tell us what you are trying to build and we will say so honestly, including when the answer is that you do not.
FAQs
How long does a discovery phase last?
The work itself is usually one to two weeks for a typical small or medium project. Elapsed time is a different question, and it is almost always decided by your side rather than ours. If the people who know how the business actually works can give us a couple of days between them, it moves quickly. If they can only spare an hour a week, it stretches, and the report is worse for it. Our own scoping step is built around that, which is why it is measured in days of your people's time rather than months of ours.
What should a discovery phase include?
Time with the people who do the work, not only the people who commissioned the project. A look at whatever systems and spreadsheets are already in use, because that is where the real process lives. Enough design to make the thing discussable rather than theoretical. And a written record of what was decided and what was deliberately left out, so nobody relitigates it halfway through the build. That record is what a fixed price is later promised against.