Redesigning the analytics dashboard merchants actually adopted

• 95% organic adoption across the merchant base, with no forced migration
• Research validated the direction with high confidence, so only low-scale layout changes were needed, which reduced rework risk going into development
• Scoped delivery so users started getting value in the next quarter rather than at the end of a full migration

Magazord merchants had data. What they did not have was a way to turn it into decisions. The existing dashboard showed numbers, but it did not answer the questions an e-commerce owner actually asks: what is working, what is falling, and what should I do this week. People opened it, looked, and closed it without acting.

The obvious path was a full migration of the module, rebuilt and shipped all at once. That plan carried two risks I did not want to accept: it would keep users waiting for value until the very end, and it would commit engineering to a large build before we had validated the direction.

I took a different position and reframed the delivery. Instead of one large migration, I proposed partitioning the work so the parts that only display data could ship first, on the APIs that already existed, delivering real value to merchants in the next quarter. Anything that wrote back to records would keep calling the legacy system in a first phase and migrate later. I brought this partitioning directly to the technology directors and aligned them on it as the plan of record.

I operated across three layers on this project rather than just producing screens.

I defined the research methodology, deciding what we needed to learn and how, so the redesign rested on evidence rather than taste. I coached the Product Manager through the data analysis, so the reads on merchant behavior were sound and the team was building its own capability instead of depending on me. And I set the success metric with the directors: organic adoption, chosen deliberately because a dashboard people are not forced to use is the honest test of whether it actually helps them decide.

The research came back with a result that is easy to undersell and worth stating clearly: the direction was right, and only low-scale layout adjustments were needed. That is not "the research changed nothing." It confirmed the redesign with a high degree of confidence, which reduced the risk of rework once development started. Validation that de-risks a build is real value, not a null result.

The redesigned dashboard reached 95% organic adoption across the merchant base. Nobody was pushed onto it. Merchants chose it, which was exactly the signal I had set out to earn.

The design work mattered, but the decision that shaped the outcome was structural, not visual. Reframing a big-bang migration into phased delivery, and taking that partitioning to the directors myself, is what let real value reach merchants a quarter earlier and kept the team from over-committing before the direction was validated. Choosing organic adoption as the metric kept us honest about whether the thing was genuinely useful, not just shipped.