They asked us to grow the one number they were already winning on.

"It works. Let's make it work more."

A founder running a healthy data-services firm came to us with a reasonable ask. The engine worked, the brand was sharp, and the instinct was the right one for a builder: take the thing that is performing and make it perform more. That instinct had a single number attached, the score the team watched and the work was graded on, and the plan was to promise customers a bigger version of it.

But the feeling underneath the ask was the one every operator knows: working harder, building more, and watching the result get eaten away anyway. The dashboard said one thing. The bank account said something quieter. The quiet worry of a founder who is good at this is not that the engine is broken. It is that they might be pouring another year into the wrong number and only finding out at the end. Because nothing looked broken, the natural conclusion was that more of the winning number would close the gap.

That belief feels safe because almost everyone shares it. Grow the metric you already lead on, promise customers more of it, scale the engine: the moves every confident team reaches for. The catch is that the easiest number to grow is rarely the one wired to revenue, so the score keeps climbing while the profit does not follow. Safe is not the same as true.

What we found when we did the arithmetic

So before touching the engine, we put two numbers next to each other on one screen: the score the firm wanted to grow, and the number that actually turned into revenue. They were not the same number. One could rise all day without the other moving an inch.

The favorite metric was a thing customers said they wanted, which made it easy to sell and easy to promise. But upstream of it, the steps that actually produced paying business were a separate motion entirely. Improving the requested number was improving a scoreboard that was not connected to the result, which is why the work could climb while the bank account sat still.

You already know this in your own work: a thermometer that reads a comfortable number tells you nothing if it is taped to the wall instead of the patient. The firm was about to spend a year cooling a room that was never the thing running the fever. We ran the goal backward to be sure, and in plain arithmetic the growth target translated into a far smaller number of new clients than anyone had assumed, reachable on a channel the firm already owned. The named gap was this: the entire planned build sat on the favorite score, and the paying number could be moved without most of it. Most of the roadmap was effort the result did not need.

The intervention

We did not touch the engine. We reconciled the two numbers, named the one that actually produced revenue as the source of truth, and rebuilt the roadmap to point at it. The reverse-math was the whole move: turn the vague growth wish into a unit count, then keep only the work that moves that count.

That meant demoting the favorite metric from headline promise to supporting evidence, and reallocating the planned build toward the small set of deals the result actually depended on. Nothing about the product changed. We fixed which number the work was being graded on, and the plan got shorter rather than bigger.

What the reconciliation showed

This is a teaching engagement, so we are careful here. The value was the diagnosis, not a scoreboard we can wave around. What changed was the shape of the plan and where the work pointed, once the two numbers were finally honest with each other.

The numbers the firm had been carrying, doubling one favorite score or chasing a much larger client count it believed it needed, were the evidence, not the point. Both pulled effort toward a scoreboard instead of toward the result. Once the paying number was named, most of the planned build read as fluff against it, and the path got shorter rather than bigger. We did not find a cleverer growth tactic. We separated the number that gets asked for from the number that pays, and the work re-pointed itself.

A note on the figures: the underlying numbers in this engagement are agency-recorded from session work and have not been independently audited, so this is shared as a teaching example rather than a results claim. The figures above describe the firm's own reasoning, not an outcome produced by us.