Finish-rate is a fitness metric

Volume is a vanity metric. Finish-rate is a fitness metric. That sentence is the whole argument, and the rest of this essay is an attempt to earn it — because the difference between the two words is the difference between a number you can buy and a number you have to be good enough to produce.

A vanity metric flatters you. It goes up when you spend, and it goes up whether or not anything happened on the other side of the screen. Impressions, reach, publish count, follower totals — every one of them will rise if you pour enough money and enough output into the top of the funnel. That is not a slander. It is a definition. A vanity metric is any number whose growth can be purchased independent of whether the work was worth reading. You can always make it larger. That is exactly why it proves so little.

A fitness metric is different in kind, not in degree. Borrow the word from biology, carefully: fitness there does not mean strength or size or how loud an organism is. It means how diagnostic a trait is of survival — how well it predicts that the thing carrying it will still be here next season. A fitness metric, in our sense, is a number that tracks whether the work actually did its job in the one place that matters, which is inside a real person's attention. It cannot be topped up from the outside. It only moves when the thing you made was good enough to move it. That is the property we are after, and it is rarer than it sounds.

What each number can prove

Start with what a metric is for. A metric exists to answer a question you cannot answer by looking. The useful question is never "how much did we publish" — you already know that; you did it. The useful question is "did any of it land." Volume cannot answer that question, because volume is measured before the reader is involved. It counts your effort, not their response. It is a record of intent.

We wrote about the mechanism a few weeks ago: volume metrics select for content that is forgettable on purpose. When the number you optimize is output, you will unconsciously optimize for the cheapest possible unit of output, because that is what lets the number climb fastest. The selection pressure runs the wrong way. Nobody sets out to make forgettable work; they set out to make the counter go up, and the counter goes up fastest when each piece asks the least of both writer and reader. Volume does not just fail to measure quality. Over time it actively erodes it, because it rewards the opposite behavior.

Finish-rate measures the response instead of the effort. It asks: of the people who started, how many stayed to the end? That is a question only the reader can answer, and they answer it with the one resource that cannot be manufactured — their finite, unrecoverable attention. The internet rewards what people finish, not what brands publish. The reward — the return visit, the forward, the slow accretion of a reason to come back — attaches to the work people completed, not to the work you shipped. Completion is where those two diverge, and the divergence is the entire game.

The one number you cannot buy

Here is the property that makes finish-rate worth building a company around. Completion is the only reader-side metric that cannot be inflated by spend.

Read that precisely, because the precision is the point. Not "one of the few." The only one. Every other reader-side number has a purchase path. You can buy impressions outright. You can buy clicks — a headline built to mislead will manufacture them, and money on distribution manufactures more. You can buy opens with a subject line that promises something the body does not deliver. You can buy the start of a read. What you cannot buy is the finish, because finishing is not a decision the reader makes at the door. It is a verdict they render continuously, sentence by sentence, and the only thing that keeps them rendering it in your favor is that the next sentence was worth the last one. There is no budget line for that. We made this case for a single un-buyable number once before; finish-rate is what that number looks like when you name it.

This is why the metaphor holds. Finish-rate is diagnostic of survival in the attention market the way a trait is diagnostic of fitness — not because it is loud, but because it cannot be faked. Anything that can be faked stops being information the moment someone has a reason to fake it, and in a market with money in it, someone always does. The value of a signal is exactly its resistance to being purchased. Completion resists completely. That resistance is not a nice feature of the metric. It is the metric. A number you could buy would tell you nothing you did not already know about your own bank account.

Be careful about what this claim does and does not assert. Saying finish-rate is diagnostic of survival is not saying completion has been proven to cause retention, or loyalty, or recommendation. It is a claim about what the number resists, not a claim about a measured chain of downstream effects. The argument is narrower and sturdier than a causal one: because completion is the reader-side number you cannot fake, it is the reader-side number worth trusting. Fitness here means unfakeable, and therefore informative. It does not mean we have run the experiment on survival itself. Overreaching on that point would be the same error we spend our time diagnosing in everyone else.

What the number does to the work

A metric is not a passive readout. Whatever you measure, you eventually make more of, because the number becomes the thing people are quietly trying to move. This is the real reason the choice matters — not that the wrong number misinforms a report, but that it reshapes the work upstream of the report.

Point a team at volume and, without anyone deciding to, the work drifts toward whatever is cheapest to produce and quickest to ship. Point the same team at finish-rate and the drift reverses. Now the only way to move the number is to make the next sentence worth the last one, all the way down — because that is literally what a reader completing the piece measures. A fitness metric cannot be gamed from the outside, so the only place left to work is the inside: the argument, the pacing, the decision to cut the paragraph that was not carrying its weight. The metric stops being a scoreboard and becomes an instruction. It tells you, concretely and continuously, that the work has to be good enough to hold a person who owes you nothing and can leave at any line.

That is what it means for a number to be worth steering by. It does not just describe whether you were good. It makes being good the only available strategy for improving it.

The wrong number and the right one

We have argued for three essays now that the number everyone reports is the wrong one, and most teams already know it. That was the diagnosis. This is the answer it was setting up. If the counted number is wrong, the obvious follow-up is: fine — which number is right? And the reason the industry keeps reaching for the wrong one is not stupidity. It is availability. The wrong number is the buyable number, and the buyable number is always the easy one to move, to report, to put in the deck. Its wrongness and its convenience are the same fact seen from two sides. It is convenient because you can move it with spend, and it is wrong because you can move it with spend.

The right number is the one with the opposite profile. It is inconvenient. You cannot move it with spend; you can only move it by being better. It will sit lower than you would like, and it will not respond to the levers that make the vanity numbers jump, which is precisely the evidence that it is measuring something real. A number that only goes up when the work gets better is worth more than a number that goes up whenever you open your wallet — even when, especially when, the first number is smaller. The discomfort is the signal working.

None of this is a call to publish less for its own sake, or a claim that output does not matter. Volume is fine as a fact about the world; it is a disaster only when it is mistaken for a measure of worth. Count it if you like. Just do not steer by it. Steer by the number that steers back.

The gate we hold ourselves to

We would not ask anyone to adopt a standard we do not sit under. So, plainly, here is ours.

We measure completion, not impressions. The bar we hold is a 65% finish-rate on long-form editorial — median per piece, on a four-week rolling window, sustained across two months before it counts as met. That is the gate, not a trophy. It is the line a piece has to clear to earn its place, stated as the discipline we impose on ourselves, not as an outcome we are claiming to have already delivered. We are pre-commercial and we are not going to dress a standard up as a result. The number is a promise we make about the bar, and a bar is only worth naming if you are willing to fail it in public.

We hold that particular line for reasons that follow directly from everything above. The window is four weeks rolling because a single piece's completion is noisy and a single good week proves nothing. The requirement is two sustained months because a metric worth trusting has to survive time, not spike once. The measure is a median rather than an average because one viral outlier should not be allowed to launder a body of work that did not finish. Every one of those choices is designed to make the number harder to fake — including harder for us to fake to ourselves. A gate that is easy to clear is a vanity metric wearing a gate's clothes.

Which number, if you can only keep one

Strip the whole argument down to a decision, because eventually every measurement philosophy becomes one. If you could keep exactly one number about your own work, and lose the rest, which would you keep?

Keep the one you cannot buy. Keep the one that goes up only when the work got better and sits still when you spend. Keep the number that would embarrass you if it were low, because a number that can embarrass you is a number that is telling the truth. Volume will always be there to flatter you, and it will always be available, and it will always be beside the point. Finish-rate is harder to move, harder to report, harder to celebrate — and it is the only one of the two that knows whether you are any good.

Volume is a vanity metric. Finish-rate is a fitness metric. Pick the one that can survive being tested.