Building the Analytics & Web Optimization team

Problem

When the experimentation function began at AEO, it was effectively one person. AEO was a bricks-and-mortar-first company, and experimentation didn't yet have organizational gravity — every test was a one-off, every prioritization was a fresh battle, and there was no team to absorb growing demand.

Approach

I started by treating the role as a learning problem before a staffing one. As the solo engineer, I implemented and ran experiments across the testing platforms in play at different times — Adobe Target, Google Optimize, and Optimizely Web & Feature Experimentation — and built up the engineering practices and patterns that would later need to scale.

The political work was as important as the technical work. Experimentation only earns budget when stakeholders can see the value, so I spent significant time making outcomes visible: showing what was learned, what was earned, and how fast each loop closed. As wins compounded, the case for staffing made itself.

Outcome

• Grew Analytics & Web Optimization from 1 engineer to 20+ resources
• Earned an increased budget at a moment when other parts of the org were tightening
• Established experimentation as a first-class function within AEO's digital strategy
• Built the team that later drove $130M+ in annual revenue from experimentation alone

What this says about how I lead

Team building inside a skeptical org is mostly evidence-gathering. The team got staffed because the program had a track record — not because I asked nicely. Once that pattern was clear, the trickier work began: structuring decentralization across six project teams so the experimentation org could scale without becoming a bottleneck.