You do not have a data problem. You have a question problem.
Teams drowning in dashboards are rarely short of numbers. They are short of a decision the numbers were collected to inform.
5 September 2025 · 5 min read
'We have plenty of data, but not enough clarity' is the most common sentence I hear. It is almost never a tooling problem. The data was collected without a decision attached, so it can support any story and settle none.
Start from the decision, work backwards
Before any research, I want to know what will be decided differently depending on the answer. If nothing will, the research is entertainment and should be cancelled.
Two kinds of evidence, used for different things
Qualitative work tells you why something happens and gives you the language customers use. Quantitative work tells you how often and how much. Teams commonly ask interviews to size markets and dashboards to explain motivation, and are then surprised that both disappoint.
Clarity is a subtraction exercise
Most clarity work I do is removing measures nobody acts on, so the two or three that would change a decision become visible. That is unglamorous and it is usually the highest-leverage hour of the engagement.
Where this shows up in the work
Keep reading
- The marketing funnel is nice, but it misses a very crucial aspect: the whyThe funnel is a model of the business, not of the customer's mind. Bob Moesta's demand-side thinking shows what is happening underneath each stage, and why that changes demos, content, search and sales.
- How to actually pull off Jobs To Be Done (without losing your mind or building features nobody wants)A practical field guide to JTBD: how to run switching interviews, map the four forces, avoid the classic traps, measure real progress, and turn it all into product, marketing and sales decisions.
- Personas describe people. Jobs explain decisions.JTBD is a lens, not an identity. Used well it explains why someone switched; used badly it becomes another artefact nobody opens.