Blog · Engineering

How we run a discovery sprint

What happens in the short, focused discovery phase at the start of an engagement, and the decisions and documents it produces before any code is written.

Published June 19, 2026Reading time 2 min

Many software projects go wrong in the first two weeks, before any code exists. The goal is fuzzy, the constraints are unspoken, and the team starts building the most visible feature instead of the most important one. A short discovery sprint is how we avoid that.

What we’re trying to learn

Discovery answers a handful of questions with enough confidence to commit budget:

  • What outcome matters? Stated in business terms, with a way to measure it.
  • Who is involved? Users, decision makers, and people affected by the change.
  • What exists today? Systems, data, workflows, vendors, and the workarounds people use.
  • What are the constraints? Deadlines (filing season, harvest, renewals), budget, security and compliance obligations, and anything that must not change.
  • What is the smallest useful release? The first version that delivers real value.

How we do it

A typical discovery sprint combines a few working sessions with the people who do the work, a review of existing systems and data, and quick sketches of workflows and screens. We look at real examples, such as a spreadsheet, a document package, or a week of the inbox, because they reveal more than descriptions.

We write things down as we go. Decisions, open questions, and assumptions are recorded so nobody has to rely on memory of a call.

What you get at the end

  • A clear problem statement and success measures.
  • A map of the workflow and the systems involved.
  • A risk register, covering what could go wrong and how we’ll handle it.
  • A proposed architecture and data model, at the right level of detail.
  • A sequenced plan, starting with the smallest useful release, with estimates.

These documents are yours. You can build with us, with your own team, or with another vendor.

When discovery is the whole engagement

Sometimes discovery shows the right answer is not to build at all. An existing product may fit, an integration may close the gap, or the process may need to change first. We think that’s a good outcome. Spending a little to avoid a large mistake is the point.

Read more about how we work and our architecture and systems design service.


Written by Syntonyl team. This article is general information, not legal, tax, or financial advice.

Bring the goal. We’ll shape the technical path.

Tell us what needs to work better. We reply within one business day with a clear next step.