One Number Is a Trap: Build a Metric Tree Your Whole Team Can Aim At

Picking one number to rally the team feels clean. One north star, everyone rows in the same direction, done. Then reality shows up. Someone games the number. The business goal drifts away from what customers actually feel. Two teams optimize the same metric and step on each other. The problem is not that you picked a metric. It is that you picked only one, and left it hanging with nothing around it.

A good metric setup does two things. It shows how customer behavior connects to money. And it gives people enough context that no single number can be tortured into looking good while the business quietly rots. Here is how to build that.

One number is easy to game, so stop leaning on one

The danger with a lone north star is Goodhart's law: once a number becomes the target, people find ways to move it that have nothing to do with real value. Itamar Gilad makes the case that one metric cannot carry a whole product and business. He suggests three true-north metrics instead: one for value to the customer, one top business KPI, and one balancing health metric that catches the damage.

The balancing metric is the part leaders skip. If your north star is signups, your balancing metric might be retention or support load. That pairing keeps a team honest. It is hard to inflate signups with junk when a second number is watching the fallout.

So before you crown a single number, ask what it might quietly break, and put a metric on that too.

Make sure customer wins and business wins point the same way

A metric can be popular and still lie to you. Justin C argues that NPS does not reliably predict revenue, growth, or retention. People like it because it is one tidy score. But a high NPS that never shows up in purchases or churn is a feeling, not an outcome.

The fix is to anchor on things customers actually do. Purchases. Repeat use. Churn. Those behaviors tie back to money in a way a satisfaction survey does not.

This is the real test of your value metric: when it goes up, does the business goal go up too? If your customer number and your business number can move in opposite directions, you have two numbers pretending to be one strategy.

Draw the tree so people see how their work connects

Most teams cannot say which numbers matter or why their daily work moves the big goal. A KPI tree fixes that. Petra Wille describes building one on a whiteboard, starting from the company goal at the top and branching down into product metrics and the customer behaviors underneath.

The point is not a pretty diagram. It is that each product manager can find their branch and see what they can actually influence. A PM does not move revenue directly. They move activation, or checkout completion, or first-week retention, and the tree shows how that rolls up.

Spend an afternoon drawing it with your team. The arguments you have while drawing are the real value. If two people disagree about which branch a metric sits on, you just found a gap in your strategy.

Let the tree steer your OKRs, not the other way around

OKRs often feel like busywork because each team writes its own and nothing lines up. Ethan Garr tells how a growth team fixed this by tying every team's OKRs back to one value-to-user number. The north star became the thing OKRs ladder up to.

Once you have a metric tree, this gets easy. Each team's objective sits on a branch. Their key results are the metrics they can move on that branch. You stop debating whether an OKR matters, because you can trace it up to the goal or you cannot.

This also kills the isolated-optimization problem. When teams can see the shared tree, they notice when their win costs someone else's branch.

The deep cut

The metric tree is not really a measurement tool. It is a way to argue in public. The value shows up when your team disagrees about where a number belongs, because that disagreement is a hidden disagreement about strategy. A lone north star hides those fights. It lets everyone nod at one number while quietly believing different things about how the business wins. Build the tree, pair your metrics, and you force those beliefs into the open where you can actually settle them.

Three questions for your team

• Are we leaning on a single metric, and what would break if someone gamed it? Name the balancing metric that would catch that damage, and start tracking it.
• When our customer metric goes up, does revenue or retention actually follow? If NPS or any satisfaction score does not link to real outcomes, decide whether it stays or goes.
• Can each PM point to their branch on the tree and name what they can influence? If they cannot, your OKRs are not laddering to the goal, and it is time to draw the tree together.