Every large company I've worked with has an innovation team. Most of them are doing the same thing: sitting in a nice office, running pilots that never ship, and producing slide decks that make the board feel modern.

I don't say this to be cruel. The people in these teams are usually smart, motivated, and genuinely frustrated. They know they're stuck. They just can't figure out why.

I can. Because I've seen it from the outside, across banks, insurers, asset managers, and large enterprises in Europe and the US. The pattern is always the same.

The innovation theater cycle

Step one: the CEO goes to a conference. Sees a demo. Gets excited. Comes back and says "we need to do AI."

Step two: a team is created. Budget is allocated. A Head of Innovation is hired. They have a cool title and a small team.

Step three: the team runs a few pilots. They work with vendors, consultancies, maybe a startup or two. Some POCs look promising.

Step four: the POCs need to go to production. This is where everything dies. IT says the infrastructure isn't ready. Compliance says they need six months to evaluate risk. The business line says they didn't ask for this. The budget gets cut because last quarter was bad.

Step five: the innovation team presents their "learnings" at the next board meeting. The cycle restarts.

I've watched this happen at least twenty times. It's not a failure of talent. It's a failure of architecture.

Why it breaks

Innovation can't work as a separate function because the hardest part of innovation isn't having ideas. It's integrating them into the machine.

When you put innovation in a silo, you create a team that can explore but can't execute. They can find interesting technology, run cool pilots, even build prototypes. But they can't ship, because shipping requires the cooperation of IT, compliance, legal, business lines, and operations — none of which report to them.

The result is a team that produces demos nobody uses. And after a couple of years, the company concludes that "innovation doesn't work" and cuts the team. The real conclusion should be: "we set up innovation to fail."

What actually works

The companies I've seen succeed at innovation don't have innovation departments. They have business problems and the freedom to solve them.

The best example I can give: a mid-size European bank I've worked with didn't create an innovation lab. Instead, their Head of Wealth Management said: "My advisors spend three hours a day on admin. I want that to be thirty minutes by next quarter." Then they gave a team the authority to pick the tools, work with vendors, and deploy — with compliance embedded in the team from the start. The EU AI Act actually helps here by giving teams a clear set of requirements to build against.

It took them four months to go live. No pilot purgatory. No innovation theater. Just a problem, a mandate, and a deadline.

The uncomfortable truth

Most companies don't actually want innovation. They want the appearance of innovation. The press release. The partnership announcement. This is exactly why banks keep buying AI they never use. The booth at the conference.

Real innovation is messy. It means changing workflows. It means some people's jobs will change. It means admitting that the way things were done for twenty years might not be the best way anymore.

That's hard. It's political. And no innovation team can do it alone.

The companies that will win the next decade are the ones that stop treating innovation as a project and start treating it as an operating principle. Not "we have a team for that." Instead: "every team does that."

It sounds obvious. Almost nobody does it.