Most product teams never make the decision to begin discovery. They just begin it. A signal arrives, 23 of 40 customer calls mention dashboard widgets, a stakeholder nods, and the team is already scoping interviews. The real decision, whether this signal deserves investigation at all, happened in silence and was never recognized as a decision. That is the argument at the center of this piece, and it is a precise one: before discovery activities start, there is a prior judgment called Signal Evaluation, and skipping it is the most expensive mistake a team can make because it never shows up as a failure. It shows up as months of rigorous work aimed at the wrong target.
The author structures Signal Evaluation around three sequential filters. Signal Strength asks if the pattern is real or just loud, volume and logo size are not proxies for representativeness. Job Connection is where most teams fail without knowing it: a signal can be frequent, genuine, and strong and still point at the product rather than the customer's underlying job. The widget request passes Signal Strength easily, then fails Job Connection hard. The actual job is 'I cannot see what matters at a glance,' and more widgets may make that worse. Strategic Alignment, the third filter, asks whether solving this advances your current position, not just whether the problem is real. A valid problem you cannot execute on is a distraction.
The piece is worth reading in full for the worked example alone. Watching the widget request pass one filter, collapse on the second, and reveal an inverted conclusion is a cleaner demonstration of jobs-to-be-done thinking in practice than most dedicated explainers manage. The broader argument, that cheaper and faster building raises the stakes on target selection rather than lowering them, is stated plainly and without padding. The framing of Signal Evaluation as a named, repeatable judgment rather than an instinct is the practical contribution here.
[READ ORIGINAL →]