Customer Insight
Data, research, discovery, and UX, understanding what's worth building.
This is the area where average PMs and great ones diverge. Everyone can run a standup and write a ticket. Far fewer can look at the same dashboard, the same five user interviews, and the same competitor launch and walk out knowing what to build next and why. That is Customer Insight. In my model it is the competency area defined as "understand and fulfill customer needs," and it is the engine room of Product Sense and Judgment.
After a decade in B2C mobile, here is what I keep seeing: most product failures are not execution failures. The team shipped the thing, on time and bug-free, and nobody wanted it. That is a customer insight failure wearing a strategy or execution costume. You can out-execute your way past a lot, but you cannot out-execute building the wrong thing. This area is your insurance against that, and it is the part of the craft I most want you to get good at.
#The three competencies in this area
| Competency | What it covers | The behaviors |
|---|---|---|
| Fluency with Data | Reading the quantitative truth: funnels, retention, experiments | Feature Evaluation & A/B Testing · Actionable Insight Generation |
| Voice of the Customer | Reading the qualitative truth, why users do what they do | User Persona & Journeys Comprehension · User Insights Application |
| User Experience Design | Turning insight into usable, well-crafted product | Design Sense & Feedback · Competitive Landscape Expertise |
The pairing of the first two is the whole point. Data tells you what is happening, and research tells you why. A PM who has only one half is dangerous. The data-only PM optimizes a local maximum into the ground. The research-only PM ships heartfelt features that nobody actually uses. The job is to hold both and let them argue with each other. theScore's product principles put it as "balance game sense with data," which is qual, quant, and intuition together. I still run my teams that way.
#Insight without discovery is just opinion
The competency model tells you what good looks like. It is thinner on the practice that produces insight continuously. So I treat discovery as a first-class discipline that sits across all three competencies:
- User Research and Discovery for the methods toolkit, and when to use which.
- Continuous Discovery for Teresa Torres' case for weekly customer contact by the team building the product, not a quarterly research project.
- The Opportunity Solution Tree for Torres' way of keeping discovery tied to an outcome instead of a feature wishlist.
- Jobs To Be Done for Christensen and Ulwick's reframe, that customers "hire" products to make progress.
- User Personas and Journey Mapping, which is useful when grounded in research and theater when not.
Discovery is the cheapest risk reduction you will ever buy. A week of customer conversations costs you a week. A mis-built quarter costs you a quarter plus the team's morale. Marty Cagan frames it as de-risking value and usability before you build (see The Four Big Product Risks). Insight is how you stop being a feature factory.
#How this area escalates
At APM you pull a funnel report and run an interview your manager designed. By Senior PM you decide which questions are worth answering and own the insight that reshapes the roadmap. By Director/VP you build the research and analytics muscle of the whole org: instrumentation, a discovery cadence, and a culture that treats "we don't actually know" as a respectable answer. The competency stays the same. The altitude climbs (see The Master Competency Matrix).
#How AI is changing this area
Of the four areas, this is the one AI is reshaping fastest, because so much of it used to be slow manual labor. Tools synthesize interviews, reviews, tickets, and survey free-text in hours instead of weeks, and self-serve analytics lets anyone ask the data a question in plain English. That hands you the gathering almost for free. What it does not hand you is the framing of the right question or the judgment of which insight actually matters, and those are exactly where the value moves. I dig into this in AI in Discovery and Research and Competencies AI Commoditizes vs Elevates.
#Continue Reading
- Fluency with Data for the quantitative half: metrics, experiments, and turning numbers into action.
- Voice of the Customer for the qualitative half: research, personas, and advocating for the user.
- Continuous Discovery for Teresa Torres on making customer contact a weekly team habit.
- Product Sense and Judgment for the upstream judgment this whole area feeds.
- The PM Competency Model for how Customer Insight fits the four-area model.