I Was Pointing Your AI at the Smaller Lever

For two years I told owners to use AI to cut costs. It works. It's also the half that matters less. (Part one of three.)

For two years I said the same thing in every room.

Use AI to cut the busywork. Automate the repetitive stuff. Save time, save money. I said it at conferences. I said it in webinars. I said it across the table in mentoring sessions. And it was true.

It’s still true. But I was handing you the smaller lever.

Cost savings is the easy thing to teach. It’s safe. It’s measurable. You point to the hours saved and everyone in the room nods. Revenue is messier. It depends on your market, your offer, your timing. So I led with the safe number. Most of the industry does the same.

Here’s what changed my mind. I started paying closer attention to the people writing the big checks and reading the hard data. And the data is rough on the cost-cutting story.

Bain surveyed 951 companies this year. Most expected AI to cut costs by 11 to 20 percent. Among the ones that actually measured it, the largest group, around 40 percent, saw 10 percent or less. As Bain put it, the technology worked but the value didn’t arrive.

Why not? Two reasons. Both matter for your business.

First, saving money has a floor. You can only cut so much. Once the busywork is gone, the lever stops pulling. Your current costs are the most you could ever save. That’s the whole ceiling.

Revenue doesn’t work that way. The only limit is the size of your market. That’s a far bigger room to play in.

Second, savings get competed away. That’s the part I missed. The logic is simple. If AI lets you cut your costs, it lets your competitor cut theirs too. Then you both drop your prices to win the same deal. The savings leave your pocket and land in the customer’s. Everyone works harder for the same money.

New revenue doesn’t leak like that. A faster quote that wins a job is yours to keep. A customer who buys more because you got to them first is yours. Nobody hands that back.

So here’s the reframe I wish I’d led with.

There are really only four ways to grow a business. Get more customers. Make each one worth more. Keep them longer. Or run the whole thing with less friction.

The first three grow the top line. The fourth, less friction, is where cost savings live. That’s the entire bucket I’d been teaching. One of four.

Now look at what you’ve actually pointed AI at so far. Cleaning up invoices. Automating reminders. Drafting emails faster. Useful, all of it. And all of it sits in that fourth bucket, the capped one.

The other three have no ceiling. They’re also messier. The path is less obvious, the payoff harder to predict. So that’s not where most of us point AI first.

That was my miss. Not the tool. The aim. Same tools, pointed at the wrong target.

So next time someone shows you an AI tool, ask one question. Which of the four is this for? If the honest answer is always "less friction," you're polishing the smallest bucket and leaving the other three untouched.

Next time I’ll walk through the three buckets you’ve been ignoring — the ones with no ceiling. That’s where your next dollar is hiding.

— Sandeep Dhall. I write about how owners of real businesses make technology, AI, and growth decisions — the ones with real money on the line. Co-founder of Bridge Digital, solving real business problems with AI and custom software. Founder of SMB Capital Partners, where we buy, build, and back small businesses worth keeping.