In a recent piece, I laid out a framework called relative moat evaluation. It maps software companies onto a spectrum from structurally protected to fully exposed, based on two questions: does AI adoption make the company stronger or the infrastructure layer stronger, and are switching costs structural, contractual, or nonexistent?

That piece was the diagnosis. This one is about what comes next: moat migration: the deliberate act of moving a company from an exposed position to a protected one. If you sit on the board of a European software company, a generic “adopt AI faster” message is not enough. In some cases it strengthens the infrastructure layer’s position more than the company’s own.

If your portfolio company owns proprietary data that the infrastructure layer cannot replicate by running its own platform, and customers interact with the application directly rather than through an agent, AI is an accelerant. Every feature shipped makes the moat deeper. The board conversation is about how aggressively to press the advantage.

Wolters Kluwer shows what this looks like. It acquired Libra, rolled out an AI-powered legal workspace across nine European markets, and is turning its proprietary content into a platform. It also reveals a second opportunity: if your portfolio company is structurally protected and competitors in the same category are being commoditised in place, this is the best acquisition window in a decade. Acquiring them means getting customers at a discount and migrating them onto a protected platform. The board should be asking not just “are we moving fast enough on AI?” but “who in our category is exposed, and should we be buying them?”

Many European vertical software companies sit here, and the board needs to act while the metrics still look strong and the company still has leverage. Three paths are available.

Path one: migrate from operational value to structural value. Doing the job well is not enough. The board should push the team to own something the infrastructure layer cannot replicate: proprietary datasets that take years of domain work to accumulate, or a regulatory workflow where the output has legal standing. The goal is to own an asset or a position that cannot be replicated by running a platform.

Path two: own the AI interface. If an AI agent is starting to sit between the product and the customer, the counter is to own the intermediation, not fight it. Build the agent interface on top of the product so the customer has no reason to use a third-party agent. If the customer talks to your AI, you keep the relationship. If they talk to someone else’s, the repricing has already started.

Path three: if neither of the first two is realistic, start the M&A conversation now. Strategic acquirers who understand the infrastructure dynamic are already screening for companies with strong customer bases and weakening defensibility. The best time to sell is while retention still looks healthy. In twelve to eighteen months the repricing will be visible, and the difference between selling before versus after could be a successful exit versus a write-down.

If the company’s core function is operational, its data sits in the customer’s own systems, and agent platforms can already access it directly, the board is not dealing with a pricing problem. It is dealing with an existence problem. No amount of AI added to the existing product will save a category that is being absorbed. But that does not mean the company is dead. It means the company needs to reinvent itself into a position where AI makes it structurally stronger, not just operationally faster.

Intercom is the clearest example. Two years ago, it was a mature customer support platform, and any company with a knowledge base and an LLM could build a basic support chatbot. CEO Eoghan McCabe shifted nearly 80% of R&D toward building Fin, an AI agent that now resolves the majority of customer queries autonomously. Fin is not a wrapper. It is trained on billions of interactions with models trained on proprietary data that improve with every deployment.

Intercom migrated its moat, from a position where AI threatened the category to one where AI is the moat. The old Intercom was commoditised in place. The new one is AI-resilient. The mechanism: do not defend the old category. Leap into a position where the framework’s two questions produce different answers. And do it while your assets (customer relationships, domain expertise, team) are still worth something.

The first piece was the diagnosis. This one is the playbook. The question for your next board meeting: which quadrant are we in, which path are we on, and are we moving fast enough?

Alessia Russo is a technology investor at Insight Partners, a $90B global software investment firm, based in New York City. Originally from Turin, Italy, she holds a dual degree from Columbia University and Sciences Po. She focuses on structured equity investments in growth-stage software and technology companies across the US and Europe.

Disclaimer: The views expressed in this article are the author’s own and do not reflect the views of Insight Partners.