Refract Group2026
Product sprint
One problem, prototyped and tested in a week.
A product sprint takes one problem in your product to a working prototype, puts it in front of your users and ends with a plan for what to build next. It takes one week and costs $2,500, credited against work that starts within 60 days.
A week on one problem
A product sprint is a fixed week of work on one product problem. We build a working prototype of the flow, your users try it, and you leave with a plan for the next build. You may know the format as a design sprint. Our studio is in Sydney, Australia, and we work between Sydney and New York with startups and software teams deciding what to build next.
One question about one user at one moment, short enough to write in a sentence.
The flow, made to click through in a browser and filled with your own content.
People who use your product try the prototype while we watch what they do.
What to build, in what order and at roughly what cost, written by people who would build it.
In short
What you get
A working prototype of one flow, a written account of what your users did with it and a plan for what to build next, in order, with a rough cost. All of it is yours, whoever builds the product.
What it costs
$2,500, fixed, for the week. If more work with us starts within 60 days, the $2,500 comes off its price in full. If you stop there, you keep the prototype and the plan.
How long it takes
One week, with a fixed end date. A sprint can often start sooner than our usual couple of weeks from the first call. If the problem is too big for a week, we narrow the problem.
When it’s not a fit
The fix is already known, the open question is pricing or positioning, the fault is technical, or you cannot reach the people the product is for. A prototype cannot answer any of those.
What happens in the week
Read what a product sprint is and when to run oneBefore the week
On a call we agree the problem and who decides on your side. You book a handful of users for the test now, because recruiting is the step that slips.
Audit
We go through what exists: the product, its analytics and its support history. The problem is narrowed to the one question the prototype has to answer.
Prototype
We build the flow so it can be clicked through in a browser, on your own content, with anything sensitive anonymised.
Test
Your users try it. We watch where they hesitate, where they stop and what they do that nobody planned for.
Plan
We write down what we saw and what to build next, in order, with a rough cost. The prototype and the plan are yours.
After the sprint
Questions about the sprint
$2,500 for one week, fixed. We agree the problem the week is for before it starts. If more work with us starts within 60 days, the $2,500 comes off the price of that work in full. If you go no further, you keep the prototype and the plan.
One week, with the end date fixed before it starts. Work usually begins within a couple of weeks of a first call, and a sprint can often start sooner than that. If the problem turns out to be too big for a week, we narrow the problem and keep the date.
Not quite. A design sprint is a defined five-day method for prototyping and testing an idea. Product sprint is a broader term. Ours means one problem, a prototype that runs in a browser on your content, a test with your users and a plan for the next build. If you came looking for a design sprint, this is what we offer in its place.
The week has four parts, always in the same order. We audit what exists and narrow the problem to one question. We build a prototype of the flow in a browser. Your users try it while we watch. Then we write the plan. How the days divide depends on the problem.
A decision-maker and someone who knows the product’s data and history, plus the users for the test. The decision-maker should be there for the audit at the start and the plan at the end. Before the week, book the users and gather content the prototype can run on, with anything sensitive anonymised. We also need to see the product, its analytics and its support history.
A working prototype of one flow, a written account of what your users did with it and a plan for what to build, in what order and at roughly what cost. All three are yours to keep. You can hand them to our team or to any other.
Run an audit when the product exists and you do not yet know where it fails. Run a sprint when you know the problem and need to learn whether a fix works before you build it. Each takes one week and costs $2,500. An audit ends with findings in order, and a sprint ends with a tested prototype.
You decide. The plan says what to build and in what order, and it is yours whether or not we build it. If you carry on with us and the work starts within 60 days, the $2,500 comes off its price. We scope the next piece and agree what it costs before it starts.
The designers and developers who would build the product, with no account manager in between. That is why the plan can carry a rough cost: it is written by people who know what the code will take. The same team designed and built Prism, our own product, from start to finish.
Yes. We run them from our Sydney studio, and the product sprint is our entry offer for product work. Most of the companies we take on have raised a seed or Series A round, and we work with earlier ones where the partnership looks like a long one.
A closer look at the work
- What is a product sprint, and when should you run one?A product sprint takes one product problem and a fixed length, tests a working prototype with the people it’s for, and ends with evidence and a build plan.
- Prototype vs MVP: what should you build first?A prototype tests a question. An MVP is a usable first product. Choose the next build by the uncertainty you need to resolve and what real use requires.
- How to brief a design agency (with a template)A good design brief states the problem, the audience, what’s fixed, what success looks like, the budget range, the deadline and who decides. Here’s a template.
About to build something nobody has tested?
Send us the problem in a sentence and who it affects. We’ll say whether a sprint would answer it.
Most teams start with a call, then a sprint or an audit.
The people you’d work with, from the first call to launch