Experimentation Hero in Focus: Erwin Vinke (a.s.r. verzekeringen)

From CRO to Technical Web Analyst

Three months ago, Erwin Vinke joined a.s.r. verzekeringen as a Technical Web Analyst, after holding similar positions at other companies. His career, however, started in CRO.
“I’ve always had a passion for optimisation, online persuasion, and psychology,” he says. “But I realised I get truly energised by technical marketing. That’s when I took the turn from CRO into a more technical role.”

Today, Erwin works on setting up and implementing measurement plans, data collection and tracking, analytics, CRO, and personalisation. His daily toolkit includes GTM, GA4, BigQuery, and now at a.s.r., Tealium.
“You could call me a data facilitator. My job is to make sure data is both available and reliable. If the data isn’t trustworthy, wrong decisions are made faster, and that’s something you want to avoid.”

A Technical Superpower

Erwin describes his superpower as the technical side of experimentation. He builds tagging and tracking setups, sometimes contributes to test builds in VWO, and writes SQL queries for reports and dashboards.
“At a previous employer, we built an automated A/B test dashboard in BigQuery using raw GA4 data. It made it so much easier for CRO specialists to monitor and analyse their tests.”

Not every test produces a win and that’s okay. Erwin recalls a series of tests for a website targeting older users with incontinence products. “We thought we were making the content easier to understand, but test after test came back negative. Sometimes you learn the most from what doesn’t work. It shows how important it is to think through your hypotheses.”

Challenges Beyond the Tech

For Erwin, the biggest challenges in experimentation often lie outside the technical realm.
“The success of CRO depends heavily on organisational commitment. If there’s no UX capacity or if development sees experimentation as a delay, the CRO train struggles to move. You need to involve departments, be transparent, and show the value of a ‘no effect’ outcome, not just the big uplifts.”

Lessons That Changed the Game

For Erwin, one of the biggest game changers in his work has been learning to work with GA4 data in BigQuery for A/B test analysis.

“If you want to analyse A/B test results sharply, I think it’s a must to work with GA4 data in BigQuery. Either yourself or with a colleague who can. It lets you create very refined analyses, with precise segments and the ability to exclude certain users. You can go much deeper than you ever could within GA4 itself.”

Another key learning: base your A/B tests on users rather than sessions, and use a maximum time window for conversions.

“In the end, you’re dealing with a user, a real person, not a session. In BigQuery, you can set rules like: the user must convert within one or two days of seeing the test, if that’s relevant to what you’re testing. That gives you much more reliable results.”

This approach addresses a common issue: visitors often see a test in their first session, then return later – perhaps hours or even days afterwards – to complete the purchase. If your analysis is session-based, you often miss linking that test exposure to the eventual conversion.

“When you control the timeframe yourself, you get results that are closer to reality. The closer to the truth, the better.”

At the same time, Erwin recognises the limitations in today’s privacy-first world:

“You’re never going to be able to trace data perfectly back to a single user. A cookie is tied to one device, after all. But you can at least get a more realistic picture. Analysts sometimes strive for this utopian 100% truth, but with all the constraints we face today, that’s simply not possible. Still, you should do your best to get as close as possible – while weighing the effort it takes and the costs involved.”

For Erwin, this mindset – aiming for precision without falling into perfectionism – is  essential to delivering trustworthy, actionable insights for experimentation.

AI as a Daily Assistant

AI already plays a big role in Erwin’s work.

“I use it to generate code for tracking scripts and test builds, and to improve SQL queries. In the next decade, much of the technical work might be taken over entirely by AI, but that will just shift my role towards prompting and directing. The key is to evolve with it.”

Advice for Future Experimentation Heroes

Erwin’s advice is twofold: keep developing yourself and understand the value of a technical web analyst in CRO.

“Keep developing your knowledge and skills and don’t get discouraged by setbacks, technical challenges, or resistant stakeholders. Find common ground and keep going. And as a technical web analyst, make sure your data quality and definitions are consistent. Your colleagues need to trust your output blindly, because bad data leads to bad decisions. That principle keeps me sharp every single day.”

Robert de Kok

CRO Specialist bij Delta Fiber