There's a story I came across recently about Daniel Humm, the chef behind Eleven Madison Park, that's been rattling around in my head. The version I heard goes like this: he took his entire staff out to one of the best restaurants in the world. The expected lesson would have been "look at what they're doing well, and let's borrow from it." That's what almost every team does on a benchmarking trip. Instead, Humm pointed his team at the opposite question. Where is even this place falling short? What's a notch off? Where's the friction? They went home and worked on those things.
It's the inversion of how most teams approach competitive learning, and it's a much sharper tool. If even the best in your category is dropping the ball on something, you've found an industry-wide blind spot — the kind of gap that's defensible to fill, because no one else is looking at it. Copying highlights makes you a slower-moving version of the leader. Closing their gaps makes you the only player solving a problem the whole category is ignoring.
I've been thinking about why that idea felt so familiar, and I realised it's because I have my own version of it. Mine just happens to involve a bank.
The Round-Up
Years ago, I was opening a new bank account. The market was — and still is — a sea of commoditized offerings. Every bank had a free checking account. Every bank had a debit card. Every bank had an ATM network. Every bank had an app that worked well enough. On the dimensions banks actually compete on, they were nearly indistinguishable. I could have flipped a coin.
I went with Bank of America. Not because of any of the headline features. I went with them because of Keep the Change — a program where every debit card transaction gets rounded up to the nearest dollar, and the difference is automatically transferred into your savings account. Buy a coffee for $3.40, and $0.60 quietly moves into savings. That's the whole feature.
It had nothing to do with the core service. It wasn't faster ATMs or a better app or higher interest. It was a small, almost cosmetic mechanic. But it worked on me — and apparently on a lot of other people too.
The numbers on Keep the Change are actually striking once you look them up. The program launched in 2005 and pulled in 2 million customers in under a year. Over its lifetime, more than 12 million customers have signed up and saved over $2 billion. At its peak, 60% of new Bank of America customers enrolled in the program, and 99% of customers who signed up stayed enrolled. That last number is the one that stops me. Bank of America found a feature that, once people tried it, they essentially never turned off. In a category where switching costs are otherwise low and customer loyalty is mostly inertia, they engineered a wedge that locked people in emotionally rather than contractually.
What's even more interesting is that the program apparently came out of a design exercise with IDEO — 80 product concepts whittled down from 20 brainstorming sessions. The bank wasn't trying to win on rates or features. They were trying to find the small thing that nobody else was doing. That's the same instinct Humm sent his team to find. Different industry, same playbook.
What Made Keep the Change Actually Work
It would be easy to read this as "add a clever feature and you'll win," which is the wrong lesson. The reason Keep the Change worked is that it satisfied three conditions at once. It was emotionally resonant — it converted a daily moment of guilt about not saving into a tiny moment of satisfaction. It was operationally cheap — a small piece of transaction logic running on infrastructure the bank already had. And it was hard to copy without looking like a follower, because once Bank of America owned the round-up mechanic in the public mind, the second bank to launch one was just trailing.
When all three hold, you get the asymmetry: a feature that costs almost nothing to build, sits on a different axis from the commodity features everyone competes on, and locks in an emotional preference that's hard to dislodge.
The Necessary Caveat
I want to be careful about one thing. Wedges are leverage on top of a working core. They are not a substitute for one.
If your core product is broken, no amount of clever round-up programs will save you. Bank of America's Keep the Change worked because the rest of their banking experience was at parity with the market. The free checking worked. The card worked. The app worked. The wedge wasn't compensating for failure on the core — it was compounding on top of competence. The story is "build a fine product and add the small thing nobody else has," not "skip the product and just do the small thing."
So the order is: get the core right, then go looking for the gaps. Both halves matter. Skipping either one collapses the strategy.
What Humm Was Really Teaching
The thing I keep coming back to about the Eleven Madison Park story isn't the tactic. It's the discipline. Most teams, when they see something great, default to admiration. Humm trained his team to default to interrogation. Excellence is interesting because of where it stops, not because of where it succeeds. If you can train yourself to look for that — at competitors, at category leaders, at the products you most admire — you start finding wedges everywhere.
The other place wedges hide is in the spaces between the features — not in survey responses, but in the small, unarticulated wishes customers don't bother sending you because they assume nothing will be done. "I always forget to do X." "I wish this remembered me." Category leaders are too busy on the headline roadmap to notice. That's the opportunity.
The next time your team comes back from a competitor demo or a benchmarking trip, try the inversion. Don't ask what they did well. Ask what they almost did, what they nearly did, what they did adequately but not lovingly. That's the round-up program waiting to be built.
Small things are not small. They're just cheaper.