The three types of experimenters and the industry conundrum holding them back

Datadog just acquired Eppo for $220 million.

If you work in experimentation, that number should make you pause. Not because you missed out, but because it’s a moment of clarity for a category that still hasn’t figured out who it serves. It reveals a tension that most vendors and teams don’t want to admit.

The experimentation industry has a positioning problem. And it’s holding all of us back.

Three experimenters, one fractured industry

When we say “experimenter,” we’re no longer talking about one kind of person. We’re talking about at least three.

1. Marketers who care about growth

They run A/B tests and personalizations to optimize websites, messaging, and acquisition flows. They shaped the modern experimentation category. Most tools and agencies were built for them. Even today, most experimentation is still driven by this group.

2. Product and data teams who care about growth

They test features, flows, and user journeys.

They work server-side, value statistical rigor, and focus on long-term performance metrics.

They’ve grown more influential and more visible.

3. DevOps and platform teams who care about reliability

They are testing infrastructure—caching strategies, deployment rules, and network performance.

Their goal is system uptime and cost efficiency, not conversion rates.

They’re quiet. But they have influence. Datadog knows it.

The conundrum: too small and too fragmented

This leads us to the core contradiction the industry hasn’t solved.

Experimentation is too small to serve only one team. But too fragmented to serve all of them well. To grow, vendors expand their reach. Feature flag tools release visual editors. Web testing tools add SDKs.

Everyone is trying to serve everyone.

But marketing, product, and DevOps each operate with different workflows, goals, and expectations. What works for one frustrates the others.

The result? Bloat. Confusion. Compromised product vision. And marketers—who still run most of the experiments—get deprioritized.

Most experimentation still lives with marketing, but the market isn’t built for them. According to the Experimentation-led Growth Report 2025 by Kameleoon, companies that run experiments across web, product, and backend systems are 75 percent more likely to grow significantly than those who test in silos.

Unified experimentation works. It drives growth. It creates alignment.

And yet, marketing remains the most siloed part of most experimentation programs.

They are not in the pitch deck. They are not on the roadmap.

But they are still testing, still shipping, still proving the model.

So where do we go from here?

The Eppo sale doesn’t just confirm that experimentation is strategic.

It confirms that the category is drifting away from the people who have made it work day to day.

If you’re a marketer, this might feel like a loss.

If you’re in product or engineering, it might feel like overdue recognition.

Either way, it marks a shift.

The future of experimentation is not one team doing more.

It’s all teams doing it better. Together.

That requires cultural change, not just better tooling.

It means seeing experimentation as a shared practice that’s built on trust, supported by data, and designed for growth—regardless of who runs the test.

Marketers may not be on the pitch deck. But they should be in the foundation.