keymodels
Customer-first product metrics
Which measures show who the customer is, what problem exists, whether the solution works, whether it is adopted, and whether outcomes improve?
Contents
A five-question framework for building a connected product measurement story from customer context to durable outcomes.
Purpose
Use this framework to build a measurement story that begins with the customer and ends with outcomes: Who is the customer, what problem exists, does the solution work, is it adopted, and does it improve the intended outcome without unacceptable harm? It prevents a product team from treating shipping, activity, or aggregate usage as proof of value.
Application
Use it when defining a product metric tree, reviewing a roadmap bet, preparing a launch, or diagnosing a disappointing outcome. Apply the five questions to one customer job or decision boundary at a time. Segment results where aggregate averages would hide materially different experiences.
What it is
The framework links customer context, problem evidence, solution quality, adoption, and outcomes. Each layer has a distinct question and evidence type. The causal story remains a hypothesis: movement in an upstream measure can support interpretation, but it does not automatically prove that the product caused the downstream outcome.
Elements
Five questions
- Who is the customer? Define the people, organization, decision unit, context, eligibility, and meaningful segments. Avoid an abstract “average user.”
- What problem exists? Measure frequency, severity, current workaround, cost, failed outcome, and confidence in the evidence.
- Does the solution work? Observe task success, quality, reliability, comprehension, accessibility, and time or effort for the intended job.
- Is the solution adopted? Track eligible exposure, activation, repeated use, breadth, retention, and abandonment at the behavior level that represents real use.
- Do outcomes improve? Measure the customer or business result, distribution of benefit, durability, and counterweights for risk, trust, cost, or unintended effects.
Metric examples
| Question | Example signal | Counterweight or context |
|---|---|---|
| Who | Eligible customers in each priority segment | Coverage and missing segment data |
| Problem | Share of observed tasks affected by the problem | Severity and current workaround |
| Solution | Successful attempts without unplanned help | Errors, accessibility, and time on task |
| Adoption | Eligible users repeating the value-producing behavior | Forced use, abandonment, and support contacts |
| Outcome | Improvement in the intended customer result | Distribution, cost, trust, and long-term effects |
Use examples only after defining the actual customer job and decision. A convenient metric is not necessarily a useful signal.
Evidence chain
For each measure, record the linked question, definition, population, period, source, owner, freshness threshold, known limitation, and decision it informs. Connect upstream and downstream measures with an explicit hypothesis, such as: “If task success improves for eligible new users, activation should rise without increasing support contacts.”
Leading/lagging links
Treat problem, solution, and adoption signals as earlier evidence; treat durable customer and business outcomes as later evidence. Test whether the relationship holds by segment and period. Do not relabel an activity metric as “leading” unless there is a credible mechanism and evidence linking it to the outcome.
Review prompts
- Did the eligible customer or problem definition change?
- Which metric moved, for whom, and compared with what baseline?
- Does the evidence support the proposed link, or only correlation?
- Which counterweight changed at the same time?
- What decision follows, and what result would cause the team to revisit it?
See Customer-first product metrics as one connected model.
Select an element to read the article’s supporting explanation while keeping the complete set visible.
Five questions
Who is the customer?
Continue your preview
Read more of Customer-first product metrics.
Create a free account to continue this advanced article preview. Complete access is included in Pro and Team, so you can see the value before deciding to upgrade.