There's a pattern I keep seeing play out right now. Leadership gets excited about AI. Someone brings it up in a meeting. A vendor does an impressive demo. Weeks later there's a budget — and a rollout — before anyone's stopped to ask the obvious question: what problem are we actually trying to solve?

Alex Bratton has seen this before, many times. He founded Lex Tech and has been watching technology rollouts succeed and fail for over 25 years. When I talked with him recently, he didn't sugarcoat it: "The technology was never the problem. It's almost always the problem definition."

Why AI Projects Fail

Think about the last time your organization adopted a new tool. Did it start with a clear articulation of the pain you were trying to eliminate? Or did it start with someone seeing a demo and getting excited?

Usually the latter. And that's where things fall apart.

When you start with the technology, you end up inventing a problem to fit it. The result is a rollout that technically works and practically goes unused, because it doesn't match how people actually do their jobs.

Alex calls this "the tool trap." AI has made it worse. Because it's so broadly capable and so heavily marketed, it's easy to convince yourself it's the answer before you've defined the question. I've seen this happen fast — sometimes within a single meeting.

The "Corporate Marriage Counseling" Problem

One of the more honest things Alex said: he often ends up in rooms doing what he calls "corporate marriage counseling." Sales, operations, IT, and leadership — all in the same meeting, all with a completely different picture of what the problem is.

Sales has one answer. IT has another. Leadership wants a number they can put on a slide. And the person who's actually going to use the tool? Often not in the room at all.

If you implement AI to serve one group without understanding the others, you've solved one problem and handed three more to someone else to deal with.

The fix is annoying in its simplicity: slow down. Talk to people before the kickoff meeting. Ask what friction they're actually living with. Find out what "better" looks like to the person doing the job, not the person who approved the budget.

Where AI Actually Earns Its Place

The companies getting real results aren't necessarily the ones with the biggest AI budgets. They found specific spots in their operations — not strategic themes, but actual moments — where people were losing time, making avoidable errors, or doing work that genuinely didn't require a human. That's where AI is useful.

Start with a simple question: where does this process feel like dragging through mud? That'll get you further than any vendor pitch.

The pressure to adopt AI isn't going anywhere. But moving fast without knowing what you're solving for isn't a strategy — it's just expensive. The companies that come out of this era looking smart will mostly be the ones that asked better questions at the start.

What's the biggest friction point your team is dealing with right now? Genuinely curious what you're working through.