Post more is the worst advice in marketing now

In April 2025, Ahrefs ran 900,000 fresh English web pages through their own AI detector and found that 74.2% of them already carried AI-generated content. Only a quarter were purely human. The interesting number sits underneath that headline though: barely 2.5% were pure machine, and almost everything else, 71.7%, was the mixed kind, a person and a model taking turns. The web did not fill up with robot spam. It filled up with people who can now produce four times what they used to, all of them following the same advice they have followed for a decade.

Post more. Be consistent. Feed the algorithm. Show up every day. That advice was sane when making a thing was the expensive part. You wrote a blog post in an afternoon, shot a video over a weekend, and the bottleneck was always the same: how much could one team physically produce. So the operators who won were the ones who produced more of it, and "post more" was just a description of how to win, dressed up as a maxim.

Then the bottleneck moved. A model writes the post in nine seconds, drafts the email, cuts the captions, spins up ten variants before you have finished your coffee. The thing that used to cost you a day now costs you nothing, and the moment the cost of making something falls to zero, making more of it stops being an advantage. Everyone has the same machine. Your competitor down the road is also producing four times what they used to, off the same subscription you pay for. "Post more" no longer describes how you win. It describes what everyone is already doing, all at once, into the same feed.

More is now the thing that hurts you

Here is the part the volume advocates skip. The output went up and the results did not follow it up the curve. In a 2025 survey by MarketingProfs and Storyblok, only 6% of B2B marketers said AI tools had significantly improved how their content actually performed. Six percent. Everyone got faster, almost nobody got better, and a lot of them mistook the first thing for the second.

Meanwhile the place all that content lands is shrinking. Seer Interactive tracked click-through rates on Google falling as AI overviews ate the top of the page, from 2.74% down to 1.62% on plain queries between June 2024 and September 2025. So the supply of content is multiplying and the demand for any single piece of it is contracting at the same time. You are shouting louder into a room that is paying less attention, and you are paying, in time and attention and the slow erosion of your own taste, for the privilege.

A reader's attention has not expanded to meet the new supply. It is the one input nobody has figured out how to manufacture more of. When you publish a fourth mediocre post this week because a tool made it cheap, you are spending down the small amount of patience your audience had left for you, and patience does not refill at the rate the machine produces. By the eleventh email this month, the open rate has told you what they think of the volume. The unsubscribe link is just the part you can see.

The scarce thing was never the making

Watch what a good operator actually does with the new tools, and it is almost never produce more. It is produce less, on purpose, with the time the machine handed back. The afternoon that used to go on writing the post now goes on the harder question the post was meant to answer: what does this audience need to hear that nobody else is saying, and is this one of the few things worth their attention at all.

That question has no AI button. The model will draft anything you point it at, instantly, which is exactly why pointing it well is what you are now paid for. It cannot tell you that three of your ten planned posts are saying the same forgettable thing, that the fourth is a genuinely useful idea buried under a boring headline, and that the other six should not exist. Knowing which few things are worth making, and killing the rest before they go out, is judgement, and judgement is the one input that did not get cheaper this year. If anything it got rarer, because the noise it has to cut through got louder.

The brands worth following understood this before the tools arrived. You followed them because every so often they made something that was actually worth your time, and you trusted that the next thing would be too. That trust is built by restraint, by the things they chose not to send you as much as the things they did. None of that changed when the cost of making fell to zero. The only thing that changed is that restraint, once a luxury, is now the whole game.

So when someone tells you the answer to a flat quarter is to post more, ask them what they would cut to make room for one thing that mattered. The advice that built the last decade is the advice quietly sinking this one. The machine already won the race to make more. The work left for you is the work it cannot do: deciding what is worth making, and having the nerve to make only that.