The Wrong Thermometer

Assets up nine percent on the year. Every internal figure in the pack is green, and not one of them is wrong. Now the number that is not in the pack: the market itself grew fourteen percent. Nine against fourteen. Nothing in those two facts is a bad quarter, and nobody will be asked to explain either of them — but that gap is a share of the market walking away, quietly, every year it persists. And underneath it sits a harder question that almost nobody asks. Of that nine percent, how much was money somebody actually won, and how much was simply markets rising? A book can grow all year and take in nothing at all. Which means this business may be growing and shrinking at the same time, and only one of those two facts is in the pack.

That is the shape of it, and it is why this kind of decline is so hard to catch. Every internal measure compares you to yourself — to last year, to plan, to the version of the business you already were. Measured that way the water is nearly always the right temperature, because you are the one holding the thermometer. The external measure asks a different question: not did we improve, but did we improve faster than the people we are competing with for the same growth. Those two questions have different answers far more often than anyone expects, and only one of them is graded by the outside world.

So here is the discipline, and run it while everything is green. Every internal measure you rely on gets a twin: the same question asked against the field rather than against your own past. Growth against the market’s growth. Your share of what is new against the share you already hold. Productivity against the people competing for the same customers, not against your own last quarter. The pairing is the instrument — a single reading tells you whether you improved, and only the pair tells you whether you are winning. And the week to build them is a week when nothing is wrong, because a red number recruits attention on its own and a green one never will.

One more pair, and it is the one people skip, because it is not a number. Your satisfaction score has an external twin as well, and it is a customer speaking unfiltered — not a summary of one, and not once. Go through whoever owns that relationship rather than around them, go back several times a year, and ask the same questions each time, so what you end up holding is a series rather than a story. One vivid conversation plus a decisive temperament is how organisations reorganise around an anecdote. The feedback that matters thins silently: nothing breaks when you stop listening, which is why a quiet room and a comfortable number are so often the same problem.

One constraint worth naming, because it is the difference between this working and it costing you. If the number you land on is not yours to investigate, do not announce that the reporting is broken — telling a finance team their instrument is a toy buys you defensiveness, or a nine-month dashboard rebuild that produces a worse pack. Take one number, one cut, to the person who owns it, as a question you cannot answer yourself. Smaller act. It is the one that works.