By Teresa Torres · producttalk.org · @ttorres on X · LinkedIn
This podcast episode between Teresa Torres and Petra Wille argues that product teams are flooded with data but that not all evidence carries the same weight. Low-effort signals like support tickets and app reviews are abundant yet rarely reveal what to build, while rich, story-based customer interviews provide the context needed for decisions. Teresa draws on interview transcripts from her work at Vistaly to show that many teams mistake demos, stakeholder meetings, and opinion sessions for genuine customer research. The conversation stresses that the goal is improving evidence quality over time rather than achieving perfection, and that coaching should encourage talking to customers rather than discourage it.
01Key takeaways
- Low-effort signals such as support tickets and app reviews show where to look but rarely reveal what to build.
- Climbing the ladder of evidence means investing more effort in story-based interviews that capture what customers were trying to do.
- Demos, stakeholder meetings, and opinion-gathering sessions are often mistaken for customer research; check the transcript to tell them apart.
- Communicate signal strength honestly so teams know how much weight a piece of evidence deserves.
- Even a mediocre interview beats no interview, so coach teams to improve rather than discouraging them from talking to customers.
- Triangulate quantitative data, organizational signal, and qualitative insight before committing to a product decision.
02Key sections
- Too much data, too little evidence
- Teams are overwhelmed by analytics, tickets, and feedback yet still make uninformed choices. The hosts question whether direct customer conversations remain necessary in that environment.
- The ladder of evidence
- Teresa's framework ranks evidence by effort and value: cheap signals are plentiful but weak, while deliberate, story-based interviews are harder to gather and far more useful for decisions.
- What transcripts reveal about interview quality
- Reviewing real interviews shows a spectrum from story-based conversations to direct-question and preference-gathering sessions, with many so-called interviews being demos or internal meetings.
- Coaching teams toward stronger evidence
- Imperfect interviews are treated as coaching moments, and Henrik Kniberg's triangle of quantitative data, organizational signal, and qualitative insight helps teams triangulate before acting.
- Skills matter more with easy software
- Because it is easier than ever to ship mediocre software, especially with vibe-coded tools, the hosts argue that interviewing skills need to improve.
03From the post
“Listen to this episode on: Spotify | Apple Podcasts You've got behavioral analytics, support tickets, sales call notes, and a feedback inbox that never empties. With all that data coming in, do product teams actually still need to go out and talk to users? In this episode, Petra Wille”
04Frameworks mentioned
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.