The announcement goes well. The energy in the room is good. A few people ask smart questions, a few people nod along, and everyone leaves the all-hands knowing that AI is coming. Then nothing happens. Or worse: a few people try it, hit a wall, and quietly decide it doesn’t work for them. Within a month, you’re back where you started, only now there’s a low hum of skepticism that wasn’t there before.
This is the most common AI rollout I see in small businesses. Not dramatic failure. Just a slow fade into indifference. And the reason is almost always the same: the introduction was made to everyone at once, and no one knew who was supposed to go first.
The all-hands announcement is a trap
There’s a logic to announcing AI company-wide. It feels inclusive. It signals that you’re not hiding anything. It gives everyone the same starting point. But it also asks every person on your team to simultaneously become curious, competent, and comfortable with something unfamiliar — while still doing their actual jobs.
That’s not inclusion. That’s chaos with good intentions.
The businesses I’ve seen do this well don’t start with a big announcement. They start with a small experiment. One task. One person. And they let the results speak before they ever schedule the all-hands.
The pattern that actually works
The first two weeks aren’t about AI at all. They’re about picking the right task. You’re looking for something your team finds genuinely tedious — a report that gets written the same way every month, a client email that’s 80% identical every time, a meeting summary that always takes longer than it should. The task needs to be specific, repeatable, and low-stakes if it goes wrong. This is not where you put your most complex workflow. You’re building a proof of concept, not redesigning the business.
In weeks three and four, you bring in one person. Not your most tech-enthusiastic hire, necessarily — your most trusted one. The person other team members already turn to. You build the first version of the AI-assisted workflow with them, not for them. They become the lead user. They have opinions. They shape how it works. By the end of this phase, they’ve saved themselves a few hours and they own the process.
Weeks five through eight are where the change actually happens. Your lead user teaches two colleagues. Not you — them. This matters more than it sounds. When a peer explains a new tool, it lands differently than when a founder explains it. There’s no performance of enthusiasm required, no sense that the boss needs this to succeed. It’s just someone they trust saying: this is actually easier, let me show you.
By weeks nine through twelve, you’re not rolling out AI anymore. You’re tidying up something that already works. You’re answering the edge-case questions, refining the process where it’s rough, and watching the team train itself. The tool has stopped being a thing you’re asking them to adopt and become a thing they already do.
What to do with the people who are nervous
There will be someone on your team who is quietly worried. Not about the task. About what AI means for their job, their relevance, their future. This fear is real and it doesn’t respond well to reassurance from a founder with an obvious interest in AI going well.
What it does respond to: being the lead user. Being the person who shapes how the tool works, whose judgment gets built into the process. When someone is worried that AI will replace them, giving them authorship over how AI is used is the most direct response available. It turns a passive threat into an active role.
This is worth thinking about before you pick your lead user. The right choice isn’t always the person most excited about technology. Sometimes it’s the person with the most to lose, given the most to gain.
The one thing to decide before week one
Before any of this starts, you need to make one clear decision: what does success look like at the end of ninety days? Not for the AI. For the person using it.
If your answer is “the team is using it,” that’s too vague to build toward. If your answer is “Sarah saves four hours a week on the monthly report and has shown two other people how,” you have something to aim at, something to measure, and something to celebrate when it happens.
The ninety days only work if you know what you’re building toward. Otherwise, you end up right back at the all-hands announcement, wondering why it didn’t stick.
If you’re not sure which task to start with, that’s actually the right place to begin — and it’s a conversation worth having with someone who’s mapped this process across a few different kinds of businesses before.
