The recurring ways good discovery efforts fail — not from bad research, but from how research is organised, shared, and acted on. The process-level traps, and their antidotes.
▸ Try the interactive toolDiscovery Anti-Patterns are the recurring ways well-intentioned discovery efforts fail — not because the research itself is poorly conducted, but because of how research is organised, shared, and acted upon. Where the previous tool catalogued the cognitive biases that distort an individual researcher, this one catalogues the process-level failures that distort a team's discovery.
These traps are insidious because the research can be perfectly good while the discovery still fails. “Discovery theatre” runs real studies designed only to rubber-stamp a decision already made; the “one-user trap” lets a single vivid story override the broader pattern; “feature fishing” asks users to design solutions instead of revealing problems; research without decision criteria produces findings that change nothing; and insight hoarding leaves good research unread in a folder. Each has a clear antidote — and recognising the pattern is most of the fix.
The research can be well-conducted and the discovery still fails — because the failure is in how research is framed, prioritised, shared, or acted on, not in the method. Good technique doesn't protect against a bad process.
A team can do genuinely competent research and still get nothing from it, because the failure isn't in the studies — it's in the process around them.
| The shortcut | What it costs | What it gives you instead |
|---|---|---|
| Discovery theatre | Research that only ever confirms the pre-made decision. | Pre-set decision criteria let research genuinely say no. |
| The one-user trap | A single compelling story hijacks the roadmap. | Weighting against the broader pattern keeps anecdotes in proportion. |
| Feature fishing | Asking what features users want, not their problems. | Researching problems gets at real needs; users design poorly. |
| Insight hoarding | Good research nobody but the PM ever sees. | Sharing and joint synthesis turn insight into team action. |
Each is a process-level failure with a specific antidote — the research can be sound while the discovery still fails on how it's framed, shared, or acted on.
| Anti-pattern | What it looks like | The antidote |
|---|---|---|
| Discovery theatre | Research designed to validate a decision already made | Pre-commit to decision criteria; allow research to say no |
| The one-user trap | A single vivid story (often a big customer) drives the roadmap | Weight it against the broader pattern; one user is an anecdote |
| Feature fishing | Asking users which features they want, not their problems | Research the problem; users are poor solution designers |
| Research without decision criteria | Running studies with no upfront definition of what would change | Define, before researching, what result triggers which decision |
| Insight hoarding | Research sits in a folder only the PM reads | Share widely; involve engineers and designers in synthesis |
A team ran a genuinely competent study — well-designed, properly sampled, cleanly analysed. And it changed nothing, because the decision had already been made before the research began. The study was discovery theatre: its real purpose was to produce a slide that said “we validated this,” not to risk a different answer.
The tell was that no possible result would have stopped the project — there were no decision criteria, so the research couldn't fail. The antidote was simple but uncomfortable: before the next study, the team pre-committed to what result would kill the idea. Suddenly the research had teeth, because it was allowed to say no — and that's the only kind of research that can actually inform a decision.
The deliverable is using this catalogue to audit your discovery process — for each effort, which anti-pattern threatens it, and is the antidote in place?
| If you see… | Apply… |
|---|---|
| Research that can't say no | Pre-committed decision criteria |
| One story driving decisions | Weighting against the pattern |
| “What features do you want?” | Problem-focused research |
| Insights only the PM reads | Shared synthesis with the team |
The hardest discovery failures to catch are the ones where the research is good. These anti-patterns operate at the process level — framing, prioritisation, sharing, action — so technical rigour offers no protection against them.
The most corrosive of the five is discovery theatre, because it wears the full costume of good practice: real studies, real users, real analysis, all in service of confirming a decision already made. The diagnostic question that exposes it — and several of the others — is simple: what result would have changed our mind? If the honest answer is “none,” the research was theatre regardless of its quality. Pre-committing to decision criteria before researching, weighting anecdotes against patterns, researching problems rather than fishing for features, and sharing insight rather than hoarding it: these antidotes all push discovery from a ritual that validates what the team already wanted toward a process that can actually tell it something it didn't.
Discovery theatre validates a foregone conclusion. Pre-commit to criteria that let it say no.
A vivid story isn't a pattern. Weight it against the broader data.
Users are poor solution designers. Research their problems, not their feature wishes.
Research only the PM reads changes nothing. Share it and synthesise jointly.
Tool 25 covers individual cognitive distortion; this covers team-level process failure. Together they cover how discovery goes wrong.
Pairing questions to decisions (Tool 24) directly prevents ‘research without decision criteria.’
Open repositories (Tool 22) and shared synthesis defeat insight hoarding.
Framing problems before solutions (Tool 04) is the antidote to feature fishing.
Recall a research effort you've seen. Ask the diagnostic question: what result would have changed the decision? If the honest answer is “nothing,” it was discovery theatre.
Then scan the other four anti-patterns — did one vivid user dominate? Was it feature-fishing? Did the insights get shared, or hoarded?
The ‘what would have changed our mind?’ test is the fastest way to catch the most dangerous anti-pattern — because theatre is research that was never allowed to fail.