The Number You Pick Becomes the Behavior You Get

Pick one number for your team to chase and something predictable happens. People start optimizing for that number. That is the whole point, and also the whole danger. If the number is a good stand-in for the value you create, great. If it is a flattering proxy, you get a team that hits the target while the business quietly rots underneath. Leaders reach for the neat single metric and skip the harder question: does this number actually capture what matters, or does it just look good in a board deck?

Here is how to choose a number you can trust, and how to catch the ones that will bite you.

Start with the value, not the scoreboard

The cleanest test for a good North Star is whether it tracks the moment your customer gets what they came for. Ward van Gasteren makes the case that a strong metric sits right at the customer's success moment, and that revenue is a bad choice because revenue is the price they pay, not the value they get. Chase revenue and you optimize for extracting money now, not for keeping people around.

The same logic kills a lot of tempting metrics. "Number of orders made" sounds like growth, but the customer gets no value from placing an order. A better version is "packages delivered without complaints," which ties the number to a happy customer, not just a transaction.

So before you lock in a number, say out loud what value your customer actually receives. Then check if your metric goes up only when that value goes up. If you can picture the number climbing while customers get angrier, you picked the wrong one.

Split the scoreboard from the plays

One North Star is not enough to run a team on. It is too broad to act on day to day. Reforge draws the line between North Star metrics and input metrics: the North Star is the scoreboard, and input metrics are the individual plays that move the score. You watch the scoreboard, but your team works the plays they can directly influence.

Reforge also gives a simple way to sharpen any candidate metric into three parts: the unit of value, the quality of that value, and the frequency. Quality is where people cut corners. Skip it and you can reward the wrong action. Their G2 example lands it: measure "visitors" and you get traffic, but measure "reviewers" and you get the content that actually drives the business.

Frequency matters too, and it cuts both ways. Push a repeat-purchase window from a month to two weeks and you may just annoy customers who would have come back on their own. Match the number to how people naturally behave.

Learn to spot the number that only looks good

A metric can climb for years while the business dies. Lloyd Tabb calls these vanity metrics: big, impressive numbers like downloads or daily active users that let outsiders compare you but do not help you run better. He tells the story of a viral recruitment platform that chased daily active users to 33 million, raised 49 million dollars, and never noticed it was losing users about as fast as it gained them. A meteor, until it fell.

His fix is clarity metrics, the operational numbers that predict behavior over time. At a call center he tested call length and extra sales before landing on something odd: attendance. Whether an agent showed up turned out to be the best proxy for accountability, and routing calls to reliable agents lifted revenue for every client.

The practical move is to look earlier than feels natural. For a car service, the vanity number is monthly riders. The clarity number is pickup time, because a one-minute wait versus ten decides whether I come back.

Ask whether you even need this metric

More numbers is not better. Stacey Barr warns about KPI overwhelm, the pile of things we track because we always have, because the data is easy, or because someone told us to. None of those reasons produce insight.

Her decision tree is worth running on any metric before it becomes a goal. Is the result written clearly enough that a 10-year-old gets it? Is it important enough to actually improve? Do you have influence over it, plus the authority, time, and budget to act? Does it fit your current strategy? If a number fails those checks, you monitor it at most, or you drop it and hand it to the team that can actually move it.

Daniel McAuley adds one more filter worth asking about any new metric: how could this be gamed? Say the answer out loud before you ship it.

The deep cut

The part that is easy to miss: a proxy is not the thing itself, and the gap between them is where your team gets lost. Nicole Gallardo uses GPS to make this stick. Follow the little blue line long enough and you stop seeing the landmarks around you, until your phone dies and you are lost a mile from home. A metric works the same way. It guides you toward a point, but it quietly stops you from noticing everything it does not measure.

That is why Bhavik Patel argues the North Star is a bad metaphor: the real star is dim, hidden, and useless if you are standing in the wrong place. A number you trust blindly stops being a guide and becomes a blindfold. Keep looking up from the line. Ask what the metric is not telling you, and go talk to a real customer to check.

Three questions for your team

• Pick our current top metric. If it doubled next quarter, could we still be losing customers or making them unhappy? If yes, we are steering by a vanity metric and need to find the clarity number underneath it.
• For each KPI we track, can we honestly say we have the influence, authority, and budget to improve it? Any number that fails this should be handed to the team that can move it, or dropped.
• How would someone game our North Star to hit the number without creating real value? Name the loophole now, then decide whether to add a quality measure or pick a different metric.