Part two of the series AI Content: 10 Strategies That Work. A new part every Friday.
Let us start by knocking down a piece of advice you will find in nine out of ten SEO handbooks: that Google rewards a regular publishing rhythm and punishes irregular bursts. Google never said anything of the sort.
What Google actually said about publishing frequency
In January 2022 John Mueller of Google was asked directly whether publishing consistency affects search rankings. His answer was, word for word, „I don’t think so“. No condition, no „it depends on your industry“, just no.
A year later he added the other side of it. In January 2023 he said that publishing frequency on its own does not make anything spam and that publishing daily is perfectly fine if you have something to say daily. That last condition is the whole joke. It is not about the rhythm, it is about whether you have something to say.
Even earlier, in 2018, he summed the same thing up from a third angle: a higher page count on a site means nothing. A thousand pages is not better than a hundred if nine hundred of them are of no use to anyone.
Publishing frequency affects exactly one thing, and that is how often Googlebot comes back to the site. That is an operational parameter of indexing, not a ranking factor. Confusing the two is like claiming that sweeping the shop floor more often improves the quality of the goods.
What Google actually penalises
Since March 2024 Google’s spam policies have included one called scaled content abuse. The definition: many pages created primarily to manipulate search rankings rather than to help users.
The key phrase in that rule is „no matter how it’s created“. The policy replaced the older rule against automatically generated content precisely so that it would apply the same way to machines and to people. Google was saying explicitly that it does not distinguish origin, it distinguishes purpose.
So no: a hundred AI articles produced overnight are not a problem because AI wrote them. They are a problem because a hundred articles with nothing to add are a hundred articles with nothing to add. The same hundred articles produced by cheap human labour would be in exactly the same position, they would just take longer and cost more.
The practical consequence for planning is uncomfortable, because nobody will give you a safe number. There is no limit that says „up to twenty articles a month you are fine“. There is only the question of whether each individual text has a reason to exist.
How to build the plan
Stop planning a number of articles and start planning what each piece of content is supposed to do. The difference shows up in the first week, because roughly half the topics usually drop out of the plan.
A practical method: every planned topic goes through three questions. What specifically will the reader be able to do or know after reading it that they could not before? What will be in that text that they cannot read anywhere else? And where will that text meet the person who needs it?
If you cannot answer the second question, do not publish the topic. Not on principle, but because a text with nothing of your own in it now competes with the answer a user gets straight from an AI summary without clicking anywhere. You lose that competition even with a well-written text.
A 70 : 20 : 10 split between evergreen content, reactions to current events and experiments is a usable guide, not a law. More important than the ratio is that the evergreen part has a revision schedule. An article about the prices of AI tools ages within three months, an article about how to brief a model lasts considerably longer. Write the date of the next check for both of them at the moment you publish.
How many articles you can actually handle
There is another way to look at publishing cadence than how much you can manage to produce. Ask how many articles a month you can genuinely check, enrich with your own data and then update six months later.
That number is your ceiling. Anything above it is debt you repay through a decline in the quality of the whole site, because content that is not updated does not stay equally good, it gradually becomes inaccurate. A text with last year’s prices does not read as an archive, it reads as carelessness.
For an editorial operation with one person working part-time, that usually works out at a handful of articles a month rather than dozens. It sounds like very little compared with what a model can generate in an afternoon. But it is the number that after a year produces a site worth linking to, instead of a pile of pages nobody maintains.
Where this usually falls apart
The most common mistake is not in the plan but in the fact that the plan survives only until the end of the first week. Not out of laziness, but because nobody is named as responsible for revisions, and what has no name attached does not get done.
Set up a simple table where every article has a publication date, a date of the last revision and the name of the person who owns it. It does not need to be a tool costing hundreds a month, a shared sheet will do.
It sounds boring. That is exactly why most people will not do it and exactly why it works. A competitive advantage in content no longer comes from a cleverer idea, it comes from someone doing the boring maintenance that everyone else gives up on after three months.
Previous part: Why most advice about AI content no longer holds. Next Friday: The prompts that make a difference, and the fact-checking without which none of it is worth anything.
The whole series: AI Content
1. Why most advice about AI content no longer holds · 2. A publishing plan that assumes volume does not work · 3. The prompts that make a difference, and the fact-checking · 4. Personalisation that is more than a different salutation · 5. Visibility in the year people stopped clicking · 6. Where AI ends and a human begins · 7. Images, and how to spot a marketing myth · 8. Measurement that tells you something · 9. One piece of content, several channels, without copy-paste · 10. Law and ethics, specifically, as they stand in August 2026 · 11. Build a system, not a pile of articles · 12. What to take from the whole series