That's not an attitude problem. That's arithmetic.
When a pilot stalls, most leadership teams reach for communication. Better framing, more training, an internal champion, a town hall about "embracing the future." All of it assumes the employee has misunderstood something.
They haven't. They've done the sum, and the sum is correct — because nobody on the leadership side has actually decided what happens to the time the AI gives back.
Here's the part that isn't printed on the card: the better your pilot performs, the worse that calculation gets. Cut forty per cent off handling time and the question lands immediately — does the team get breathing room, or does the quota move? If you haven't answered that before deployment, the person doing the work will answer it for you, based on everything they've seen happen to recovered time before. They will usually be right.
So adoption friction isn't always a people problem you need to fix. Sometimes it's a signal about terms you haven't set.
The fix is unglamorous and it costs you nothing technically. Decide, before you ship: for the first six to twelve months, where do the gains go? What share to reduced backlog, what share to quality, what share to training, and what happens to targets. Write it down. Name who is accountable for holding it. Give HR real authority over that line — not a seat at the launch comms meeting, but a say in performance expectations before the build starts.
That's not generosity. It's the difference between a commitment staff can test and a reassurance they're asked to take on faith. One of those changes the arithmetic. The other is just a nicer way of saying trust us.
Technical success and workforce benefit are two separate outcomes. Your pilot can hit every metric and still leave the people running it worse off — and those people are the ones deciding, quietly, every day, whether this thing works.
Most leaders diagnose the resistance. Very few audit the offer.
Discover more from Leverage AI for your business
Subscribe to get the latest posts sent to your email.
Previous Post
A Surprising Amount of Model Cost Is Architecture Debt