Here's a sentence I hear in every boardroom, usually in a lower voice than the rest of the meeting.
"It's not working like we wanted it to."
Nobody says it on stage. Everybody feels it. And it arrives right on schedule, about a year and a half after the big rollout.
So let me tell you what's actually happening, because it's happened before.
What's the pattern?
Every technology revolution I've studied, and I've lived through four from the inside, runs the same cycle. Panic first: the new thing reads as the end of the world. Then adoption: everyone gets the tool, mostly to do the old work a little faster. Then the dip. The results disappoint. The numbers don't move. The board asks what they paid for.
And then, if you keep going, redesign. Work and roles get rebuilt around what the new thing can actually do. That's where the returns arrive. And finally, ordinary. It disappears into the furniture. Nobody wants the payphone back.
You are in the dip. So is everyone else.
Here's the evidence. According to McKinsey, about nine in ten companies now use AI somewhere. About four in ten see any effect on earnings. And according to the Census Bureau, only four percent of adopting firms have redesigned how they work in any comprehensive way.
Nine in ten have the tool. Four in a hundred changed the work.
That gap is the dip. It's not a failure of the technology. It's a description of where you are on the map.
And here's the thing about the dip. It's the moment most organizations quit. They call the rollout a disappointment, they quietly stop funding it, and eighteen months later a competitor who kept going comes out the other side with a redesigned operation.
The dip is where the leaders get separated from the adopters.
What should leaders do now?
So, three things.
Name it. Tell your team: this is the dip, it's expected, here's what comes next. You'll feel the room exhale.
Stop measuring adoption. You won. Everyone has the tool. Start measuring redesign: how many processes actually changed shape.
And pick one thing to rebuild all the way through. Not a pilot. One workflow, redesigned around the capability, end to end. The returns show up there first, and then everyone believes you.
We've been here before. Electricity had a dip. The internet had a dip. Every one of them ended the same way, for the organizations that kept going.
You're not behind. You're on schedule.
Where's your dip right now, and who's about to quit in it?
Related: How do you get employees to adopt change they didn't ask for?
- McKinsey, The State of AI, 2025 (adoption and earnings-effect figures).
- U.S. Census Bureau, Business Trends and Outlook Survey, 2026 (share of adopting firms with comprehensive redesign).
More like this on LinkedIn. Follow @samradofficial
