Good enough, right now beats rigorous, too late. A research practice built to run at the speed agile teams actually ship at — not the speed research textbooks assume they do.
Most research methods were designed for a world with quarters, not sprints. By the time a "proper" study is scoped, recruited, run, and synthesized, the team has already shipped three iterations on gut feel. The research is thorough — and irrelevant, because the decision it was meant to inform has already been made.
Instead of a single research phase that gates the whole project, discovery and delivery run as two continuous, parallel tracks. Discovery is always working one sprint ahead — validating what delivery will build next — so research never blocks shipping, and shipping never outruns what's actually been validated.
Research is a standing part of the sprint cadence, not a phase that happens once and then stops.
A 20-minute guerrilla test today beats a three-week study that ships after the decision's already made.
Insights are framed as decisions and tickets engineering can act on, not research reports that sit unread.
It runs far enough ahead that it's never the bottleneck, and never so far ahead it's disconnected from what's actually being built.
This isn't an argument for skipping rigor — it's for matching rigor to the size of the decision. High-stakes, hard-to-reverse calls — pricing, compliance flows, anything touching sensitive data — still get the slower, deeper treatment. Speed is the default; depth is a deliberate exception, not a compromise.