You Can Buy an AI Team. Keeping It Is a Different Deal.

More and more acquisitions aren't really about the product. They're about the people who built it.

Elangovan Perumal makes the case well in his recent LinkedIn article, Buying AI Capability: When M&A Becomes a Talent Strategy. He looks at deals like Accenture's acquisition of Faculty, which he notes brought more than 400 AI-native professionals into Accenture, and argues that large companies are buying established teams instead of hiring individuals. His reason is simple:

"Hiring individual experts does not automatically create a functioning delivery organization."

The panel at a recent AI in M&A webinar landed in the same place from the buyer's side. Strategic buyers are hungry for talent that can build AI, and retention and earnouts are becoming a bigger differentiator. AI can, in theory, replicate a piece of technology. It can't instantly replicate a business: the customers, the traction and the people who know how to make it work.

So in these deals, the thesis is a bet on people. Not on whether they stay on the payroll, but on whether they keep working the way they did before the deal.

What you're actually buying

Perumal puts it precisely:

"The people, their accumulated knowledge, their ability to work together and their experience with customers form part of the Customer Value being acquired."

Their ability to work together is the part that matters most, and the part almost nobody diligences. A specialist team performs because of how it's organized: who makes which calls, how fast decisions move, how much freedom the people closest to the work have, how they handle pressure. Those habits are the capability. They're also exactly what an integration plan tends to change first.

He's honest about the risk, too:

"Acquiring a company does not guarantee retention of experts and its IP."

Where people intelligence fits, and where it doesn't

This is where AI and behavioral science earn their place. Behavioral science gives you a structured way to read how teams communicate, decide and respond to change, instead of relying on management's impressions or a handful of interviews. AI makes that practical at deal speed. It can surface patterns from behavioral data across whole teams in days, not months, and compare them with how the acquirer works, so you can see where the two will clash before the integration plan forces them together. What used to be gut feel from a site visit becomes something the deal team can test, question and plan around. The value isn't in the tool's answer. It's in the sharper questions it puts in front of the people making the decisions.

Why money doesn't protect the bet

The standard toolkit is financial. Retention bonuses for the people the buyer can't afford to lose. Earnouts for the founders. Both have a place. Neither protects what made the team worth buying.

Earnouts can even work against it. Buyers often promise autonomy to win the seller, and the earnout then gives the seller a reason to resist any change that touches their numbers. The promise and the integration plan pull in opposite directions, and the team is caught in between. Not every promise is bad, and not everything needs integrating right away. But the expectations have to be set early.

Retention money has a similar limit. It can keep someone in their seat for a year. It doesn't tell them what their role is now, who decides, or whether the way they work will survive. Clear expectations and deliverables do more than throwing money at people.

That's why the order matters: heart, then mind, then pocket. People first need a reason to stay, then clarity on what's expected of them. Money comes after. Reverse the order and you end up paying people to stay without giving them a reason to.

Interested in where AI in M&A is heading? Join our AI in M&A LinkedIn Group to exchange ideas with others working across the deal lifecycle.

Where capability actually gets lost

It's rarely one dramatic decision. It's speed and small changes. A new approval layer. A reporting line that moves. A team that used to sit together, split across two functions. Fast serial acquirers can break the strengths of smaller acquired businesses simply by moving too fast.

Incentives can bridge the transition instead of just paying for it. One approach is to hold the bonus pool back as dry powder and pay it after close, using the first six months to bridge promotion cycles. People see which behaviors get them to the next level in the new organization, and the money reinforces the clarity rather than replacing it.

Perumal's list of what deserves attention is a good one: leadership continuity, retention incentives, decision-making autonomy, career opportunities and knowledge transfer, "from the beginning of the transaction." From the beginning means diligence, not Day 1.

Planning for the team before signing

The data room only gets you so far. It tells you headcount, compensation and contracts. It doesn't tell you how a team decides, how it reacts when governance changes, or which parts of the thesis rest on a small group working a particular way. Real conversations, tangents and time on site do. So does reading culture signals consistently, and allowing for country norms in cross-border deals.

Three questions are worth answering before signing:

• Which parts of the thesis depend on this team working the way it works today?
• What autonomy have we promised, and does the integration plan honor it?
• What changes for this team on Day 1, and what can wait?

Where people intelligence fits, and where it doesn't

People intelligence can help here, if it's pointed at the right question. Used at team level before close, it shows how the acquired teams communicate, make decisions and work under pressure. That gives the integration plan something concrete to protect: which teams decide fast and informally, where decisions bottleneck, which groups will feel a new approval layer first.

What it should not do is predict which individuals will leave. Scoring named people for flight risk turns a retention problem into a surveillance problem, and it pushes the conversation straight back to money: who to pay to stay. Team-level patterns, in groups large enough that no one is singled out, are associations, not causes. They show where to look and what to ask. The judgment stays with the people running the deal.

The structural choice

In a capability deal, you're not buying a list of names. You're buying a way of working. Retention packages protect the list. Only the integration plan can protect the way of working, and it has to be written with that in mind before signing.

Pay for retention if you need to. But plan for the team.

Join the AI in M&A LinkedIn Group to exchange ideas with others working across the deal lifecycle.

This piece is part of a series running through the Under the Hood of M&A newsletter.