I built a tool that scored how well an idea would spread. A scoring tool flatters you. So I tore it down and built one that tells you the truth.

Here's what the first one did. You pasted in a pitch, a post, a paragraph, and it graded the thing against fifteen markers and handed back a number. People used it. Some got the badge. And the whole time it was quietly committing the same error I'd committed building it: it evaluated whatever you gave it without ever asking whether the idea deserved to exist in the first place. It optimized. It never questioned.

The reason traces back to a borrowed question. The scoring logic leaned on a sales-shaped model of value — dream outcome, probability, time, effort — which is a perfectly good model for one specific thing: getting a person in front of you to say yes. I had, without noticing, built a conversion machine and called it a transmission machine. Those are not the same machine. They aren't even pointed in the same direction.

Here's the difference, and it's the only idea in this essay, so I'll say it once and then build on it.

Persuasion asks: how do I make this convincing enough that someone says yes? That question ends when the transaction ends. Yes or no, the exchange closes, everyone goes home.

Transmission asks a question that only begins after that point: what does a person get for voluntarily repeating this to someone else — when nobody asked them to, and nobody is paying them?

That's not a persuasion question. It's an incentive question. And once you see it that way, the whole problem changes shape. You stop auditing your writing for convincingness and start auditing it for payoff — not your payoff, the reader's. Because a stranger who repeats your idea is doing your marketing for free, and nobody works for free. They're getting paid in something. The only question is what.

The currencies are older than any platform. Status: repeating this makes me look sharp in front of people whose opinion I care about. Identity: saying this expresses who I am. Belonging: this signals I'm one of us. Validation: this articulates something I already suspected but couldn't put into words — which might be the most powerful of them all, because you're not giving the reader a new idea, you're giving them the words for one they were already carrying. Then there's hope. And plain utility: this is a tool I'll reach for again, and I want the people around me to have it too.

Notice what's not on that list: probability the deal closes. Not one of those currencies maps onto conversion. A pitch can be maximally convincing to the person across the table and offer absolutely nothing to anyone else in the room — which is why so much competent writing dies quietly at the moment of its small success. It won the yes. It just never gave anyone a reason to carry it out the door.

None of this is new, and that's part of the point. Dawkins didn't invent ideas competing to be repeated when he coined "meme" — he named something humans had been living inside of forever. Religions, proverbs, songs, QWERTY keyboards: the ideas that survived were the ones that paid their carriers. I've written about this pattern before — how QWERTY won, how startup culture propagates itself, how two very different NFT projects understood transmission while everyone else was optimizing floor price, how ideas think you as much as you think them. This isn't a lens I'm applying to writing. It's a description of what writing has always been up against.

So the rebuilt tool does one thing the old one couldn't, and I mention it only because it's evidence the reframe holds up under contact with real material. The old version graded whatever you pasted. The new one is allowed to say: don't publish this. Not "here's your score," but "this idea, as written, offers the reader nothing for carrying it — stay quiet, or come back with something worth repeating." That single capability — the permission to reject — falls straight out of asking the incentive question first, because if the answer to "what does someone get for repeating this" is nothing, no amount of polish downstream fixes it.

Which brings the question back around, and I'd rather leave it with you than resolve it.

You're about to write something today. A post, a pitch, an email, a talk. Before you ask whether it's convincing — before you ask anything about it at all — ask what you're actually offering the person who might repeat it. What do they get? What do they get to be, in front of someone else, by handing your idea along?

If you don't have an answer, that's your answer.