Proto-Personas Presented as Research-Backed
A PM shows you a set of "personas" the team built during a workshop two years ago, where stakeholders guessed at what users needed based on gut feel. The deck describes them as "based on user research," and the roadmap for the next two quarters is being prioritized partly around what these personas supposedly want.
- Are these proto-personas or research-based personas, and what specifically distinguishes the two?
- What's the concrete risk of continuing to make roadmap decisions based on this artifact as-is?
- With a small budget (say, enough for 6-8 short interviews), what would you do in your first two weeks to validate or replace them?
- How would you raise this with the PM and stakeholders without alienating the people who built the personas in good faith?
1. Proto-persona, mislabeled as research-based
These are proto-personas: built entirely from stakeholder assumptions in a single workshop, with no interviews or behavioral data behind them. What distinguishes a proto-persona from a research-based one isn't polish or detail, it's the evidence base — a proto-persona makes the team's existing beliefs explicit so they can later be checked against real data, while a research-based persona is built from that data via pattern clustering across actual interviews or behavioral logs. Labeling this deck "based on user research" is the core problem: it launders unvalidated opinion with the authority of evidence that was never actually gathered.
2. The risk of proceeding as-is
Roadmap decisions are being prioritized around what a group of stakeholders guessed users wanted, not what users actually need — this is "persona theater" doing real damage rather than just sitting unused on a wall. The risk compounds because the team believes it's working from evidence, so contradicting data (support tickets, analytics, actual user complaints) is more likely to be dismissed as noise rather than treated as a signal the personas are wrong. Two quarters of roadmap work could ship against a persona that doesn't predict real behavior at all.
3. First two weeks with a small budget
Run 6-8 short (20-30 minute) interviews targeting a spread of actual users, focused on behavior and specific past decisions rather than opinions ("walk me through the last time you did X," not "would you like feature Y"). In parallel, pull whatever behavioral data already exists cheaply — support ticket themes, funnel analytics, sales call notes — since that's real signal that predates any new research spend. Cluster what comes back into 2-3 patterns and check each existing persona trait against that clustering: which traits hold up, which don't, and which weren't even addressed by the original workshop-based guesses.
4. Raising it without alienating stakeholders
Frame it as strengthening an existing asset rather than discrediting it: "the team did the right thing building a shared reference point early — now that we have budget for research, let's validate it against real users so the roadmap decisions built on it hold up." Avoid a framing that implies the original work was wasted; a proto-persona built honestly at the time, under the constraints that existed then, was a reasonable move — the actual issue is that it was never revisited or relabeled once it started being treated as research-backed. Bring the interview and analytics findings as concrete "here's what changed / here's what held up" rather than a blanket "this was wrong."
Share this question