One evidence base for a B2B platform across 35 markets

In one line: When the platform had more requests than budget, I combined behavioural data, surveys, and customer visits into one evidence base so leaders could compare opportunities on the same terms.

Context

A B2B platform with several billion SEK in annual sales running through it, serving dealers and distributors across 35 markets, needed prioritisation — not more features on the backlog. Product and business leaders needed a shared evidence base so they could compare opportunities on the same terms, grounded in usage and customer work rather than local anecdotes alone.

Why it was worth building

Building without a shared evidence base meant every market argued from local anecdotes. The cost of inaction was endless debate without comparable facts on the table.

What I built

• Single evidence model — 40,000 daily sessions, 1,470 survey responses, and 30 customer visits interpreted together.
• Power BI decision material — behaviour, feedback, and field context on comparable axes.
• Repeatable chain — observe usage → validate with surveys → deepen with visits → map to constraints → feed leadership comparisons on the same axes.

How it was adopted

Product and business stakeholders used the combined views when comparing what to prioritise next. Qualitative visits explained workarounds and local process; analytics showed where usage concentrated and where friction appeared.

Effect

A shared language for trade-offs across markets — not a one-off dashboard, but a repeatable way to compare opportunities on the same terms.

Where AI fits now

The same discipline applies to AI features: measure adoption and quality, not demo enthusiasm. Product analytics plus structured customer evidence is how I still ask whether an automation or agent is worth running in production.