Most customer organizations are already using AI in some form. We’re summarizing conversations, assisting agents, automating workflows, identifying risk, improving self-service, and making sense of more customer data than we could before. Much of the early focus has understandably been on productivity and efficiency.

I think that’s the first chapter of AI, not the end state.

The question I’m increasingly interested in is this: What happens when the customer organization no longer has to wait to know?

Think about how much of the traditional customer operating model is built around reacting. A customer contacts Support because something has gone wrong. Customer Success notices adoption has fallen. An implementation misses a milestone. A renewal becomes a risk. Someone eventually sees the signal, interprets it, and decides what to do.

AI has the potential to compress that entire cycle.

If I can see the conditions that typically precede churn, I can intervene earlier. If product usage, support history, sentiment, adoption, and commercial data tell me a customer is struggling, I shouldn’t need to wait for a survey or escalation to confirm it. If thousands of customer conversations reveal the same friction point, that intelligence should reach Product and Operations while there is still time to do something about it.

The real shift with AI is not from manual work to automated work. It’s from reactive customer management to anticipatory customer management.

That distinction changes how I think about the operating model.

For years, we’ve built customer organizations around functions: Implementation, Customer Success, Support, Services, Renewals. There are good reasons for those specialties, but customers don’t experience those organizational boundaries. Their data shouldn’t either.

As intelligence becomes easier to connect across the lifecycle, I think we have to question some of the handoffs and structures we’ve accepted as necessary. Why should the customer have to repeat context when they move between teams? Why should a health signal sit inside Customer Success if it should change a product decision? Why should Support repeatedly solve an issue that customer intelligence tells us should be eliminated altogether?

This is also where I think the conversation about AI and people needs more nuance. I’m less interested in how many tasks AI can remove than in what becomes possible when talented people get that capacity back.

I don’t want people spending their time searching across systems, compiling information, summarizing conversations, or repeatedly solving predictable problems. I want them interpreting what the signals mean, solving the unusual problems, building relationships, making judgment calls, and finding opportunities we didn’t know were there.

That requires us to rethink more than technology. It means questioning roles, workflows, decision rights, metrics, and even where accountability sits across the customer journey.

There’s a practical test I keep coming back to. After implementing AI, if the organization is fundamentally working the same way, we probably haven’t transformed very much.

That’s the opportunity I see now. AI gives us a chance to move from reacting to customers to anticipating them, from fragmented signals to connected intelligence, and from automating work to reconsidering why the work exists in the first place.

That is a much more interesting transformation than simply doing what we already do faster.