Saying no

December, a call, and a request that arrived at the end when everyone was already wrapping up. Could we also do the thing with the pop-up. And I said yes, and I knew within about four seconds that it was the wrong answer, and I said it anyway, and then I thought about it on and off for two weeks.

Every piece of advice I’ve ever read about this treats it as a character problem. Be assertive. Learn to say no. Protect your time. All of which is true in the way that “eat less” is true, and about as much use in the moment.

The reason I said yes wasn’t that I lacked the nerve. It’s that “no” on its own leaves the person you’re talking to holding a problem with no way forward, and they’d have pushed, and I’d have had nothing else to say. Yes was the only complete sentence I had.

No isn’t a sentence, it’s a negotiation

That’s the whole reframe, and everything below follows from it.

If somebody wants something and I say no, the conversation isn’t over, it’s stuck. They still want the thing. They still have whatever pressure made them ask. From their side, nothing has changed except that a route is closed and no other route has been offered, so the rational move is to push, and the pushing is what everybody’s actually afraid of.

A bare no leaves the other person holding the same problem with one fewer option. Of course they push. What else is there for them to do?

Which means the skill isn’t refusal. It’s having a second sentence ready.

Three different noes

I lumped these together for years and they need completely different treatment.

The capacity no. I cannot do this in the time available. This is arithmetic and it’s the easiest of the three, and I still get it wrong by making it about me (“I’m really busy”) rather than about the work.

The version that works is to make the trade visible. There are 11 days before launch, here is what’s in them, this is a three-day thing, so it goes in and something comes out. Which of these would you like to move? That isn’t a refusal at all, it’s a question with a spreadsheet behind it, and the client makes the decision, which is correct because it’s their project.

The judgment no. I think this is a bad idea. Much harder, and here’s where I’ve changed my mind over the last few years.

This one is advice, not a veto. If I explain the reasoning, and they’ve understood it, and they still want it, then it’s their business and their money and they’re allowed. The pop-up will annoy people; they may have decided the annoyance is worth it, and they know things about their commercial position that I don’t.

What I owe them is a clear statement of the risk, once, not repeated, and then good work on the thing they’ve chosen. What I owe myself is a written note of the concern, so that when it’s raised in six months there’s a record of the conversation rather than two people’s memories.

I used to think this was capitulation. I now think the developer who treats every technical opinion as a veto is not more principled, they’re just harder to work with, and they get told less.

The ethical no. I won’t do this. Faking reviews, a consent flow designed to mislead, tracking somebody has explicitly refused, an interface built to make canceling harder than subscribing.

This one is different in kind. It isn’t a negotiation and there’s no second sentence and no alternative offered, and the right way to say it is short and unembarrassed: I’m not going to build that. No lecture. The lecture is where it turns into an argument you can lose.

It should be rare. If everything is an ethical no, none of them carry weight and you’re just difficult. I’ve used it properly maybe four times in a decade and it worked all four times, which I suspect is because it’s so obviously not my usual register.

Say no to the mechanism, not the goal

The practical version of all this, and it’s the sentence I wish I’d had in December.

Almost nobody wants a pop-up. They want more people on the mailing list. The pop-up is a mechanism somebody has attached to a goal, and the goal is nearly always reasonable while the mechanism is nearly always the first thing anyone thought of.

So: never argue with the goal. Argue about the mechanism, and bring another one.

“We can definitely get more sign-ups. A pop-up will do it, and it’ll cost us on the bounce rate and it’ll be the first thing anyone sees. There’s a thing at the end of the articles that people who’ve actually read something will see, and we could try that first and measure it. If it doesn’t work in six weeks, we do the pop-up.”

That’s not a no. Nobody in the room experiences it as a no. And it’s the exact outcome I wanted, plus a mechanism for being proved wrong, which I need because I might be.

This is the same move as the brief post from a few years ago, which is that the thing you’ve been handed is somebody’s answer rather than their question, and most of the job is quietly working back to the question without making anyone feel stupid.

The bit nobody says

Here’s what’s missing from every article I’ve read on this, and it’s the reason I find most of them slightly insulting.

The ability to say no is a function of your financial position, not your character.

A freelancer with one client, whose rent depends on that invoice clearing, does not have the same options I have with a pipeline behind me. A junior on a visa tied to an employer does not have the same options as a senior with three recruiters in their inbox. Somebody supporting a family in a town with two agencies in it does not have the same options as somebody in San Francisco who could be re-employed in two weeks.

So when I read advice telling people to be more assertive, I’d want it to say that this is easier the more secure you are, and that if you can’t afford it right now that’s a fact about your circumstances rather than about your spine. Saying yes to something you disagree with, because you need the money, is not a moral failure. It’s a Tuesday.

And the thing that actually increases your capacity to refuse isn’t confidence. It’s a second client. It’s savings. It’s a network. Those are the levers, and they’re slow and unglamorous, and they’re real.

What works, mechanically

Four small things, in descending order of usefulness.

Don’t answer in the room. “Let me have a proper look and come back to you this afternoon” has never once been refused and it removes the entire social pressure that produces the bad yes. Same as with estimates, which is not a coincidence.

Put it in writing. It’s much easier to be clear in an email than in a call, the recipient can react in private without needing to save face, and there’s a record.

Be specific about what you can do. Vagueness reads as reluctance and invites another go. A concrete alternative with a date on it ends the conversation.

And keep the reasoning short. One reason, stated once. A long justification signals that you’re still arguing yourself into it, and any negotiator can hear that, and they’ll pick whichever reason looks weakest and work on it.

The pop-up, in the end

I went back on it in January, in writing, having thought about it. Offered the alternative. They took it. It took one email and about 11 minutes and I’d spent two weeks not doing it.

Which is roughly the ratio for this sort of thing, in my experience.