Engineering · Prototyping

A first version that survives the second.

Prototype and MVP builds for companies proving a market. The constraint we work to is that version two should be an extension rather than a rewrite.

  1. Scope

    Agree the one question the prototype has to answer, then cut everything that does not answer it. Most prototypes fail because they were asked three questions and answered none of them clearly.

  2. Build

    Fixed cadence, working software from the first week, in your repository. Shortcuts get taken deliberately and written down, so the debt is known rather than discovered later.

  3. Test in market

    Instrumented before launch, so what you get back is data rather than an impression of how the launch felt.

  4. Decide

    Extend, redirect or stop. All three are fine, and we will tell you which one we think it is.

What we build properly even in a prototype

Data model, integration boundaries, deployment path. Those are expensive to change later. Interface, feature depth and polish stay deliberately thin, because they are cheap to redo and you do not yet know which of them will matter.

If a prototype does not need to be built to answer the question, we will say so in discovery. It costs us a project and saves you a quarter.

Prototype and MVP builds.

Start with the question, not the feature list.

Contact us