Product Management Craft · 18 min · 120 XP

Discovery in two weeks

Find the assumption that would change your mind, and test that one, in two weeks rather than two quarters.

Discovery is not a phase before the real work. It is the habit of finding out whether the thing you're about to build is worth building, continuously, while you build other things. Teresa Torres's phrase for it is continuous discovery, and the practical version is small: every week, some contact with the people who have the problem.

Marty Cagan's framing of what can go wrong is the most useful checklist here. Any idea carries four risks: - Value: will anyone want it? - Usability: can they work out how to use it? - Feasibility: can we build it? - Business viability: does it work for us legally, commercially, operationally? Most teams test feasibility (engineers estimate it) and usability (designers test it), and simply hope about value. Value is the one that kills features.

So start by writing down what you are assuming. Not what you know, but what has to be true for this to work. “Teams want a shared inbox.” “Admins will invite their colleagues.” “Support can handle the extra volume.” Then plot each one by how risky it is against how much evidence you actually have. The assumptions that are high-risk and evidence-free are the only ones worth spending a fortnight on.

Assumption map: only the top-right corner is worth testing
                 high risk
                     │
  "Admins will       │   "Teams want shared
   invite their      │    approvals at all"     ← test this
   colleagues"       │
─────────────────────┼─────────────────────
  "Engineers can     │   "Our auth supports
   build it in       │    multi-seat billing"
   a quarter"        │
                     │
                 low risk
    no evidence ◄────┴────► strong evidence

Then run the smallest thing that could change your mind, often called a riskiest assumption test. Not a small version of the product: a test of the belief. If the assumption is “teams will pay for this”, a landing page with a real price and a real button tests it in days. If it's “admins will invite colleagues”, look at whether they invite anyone today. An MVP is a small product; this is not a product at all.

Timebox it and name the decision up front. Two weeks, one question, and (written down before you start) what result would make you stop. Discovery without a pre-committed decision rule turns into research that confirms whatever the team already wanted to build.

Do it this week: write down the five assumptions behind whatever your team is about to build, mark which have real evidence, and take the riskiest evidence-free one to someone outside the building.

Loading your workspace…