AI has changed the maths of content production. Work that used to take a week now takes an afternoon, and a lot of brands have responded the obvious way — by publishing far more. Five blogs a week where there used to be five a month. Daily social where there used to be weekly.

It feels like progress. More published, more chances to rank, more shots on goal. But the honest answer is that producing five times the content does not produce five times the result — and depending on how it is made, the extra volume can actively cost you visibility. Here is what is really going on, and what to do about it.

Volume was never the thing Google penalised — value is

There is a persistent myth that Google is “against AI content”. It is not. Its own documentation is explicit: the policy targets content made to manipulate rankings rather than help people, regardless of who or what produced it. Google describes the abuse as generating many pages for the primary purpose of manipulating Search, with little or no value added for users — and it applies whether the work is automated, human, or a mix of both.

In other words, the production method is not the problem. A thousand thin pages written by humans are treated the same as a thousand thin pages written by a model. What matters is whether each piece genuinely adds something that did not already exist.

This is why two brands can publish at the same cadence and get opposite outcomes. The one publishing genuinely useful, original material at higher frequency is fine. The one using the new speed to flood the same topics with competent-but-generic filler is exposed.

The March 2026 core update made this much less forgiving

The most recent data should give anyone scaling content pause. Following Google’s March 2026 core update, sites relying on generic AI output without human editorial oversight saw traffic fall by 60 to 80 percent. Sites that used AI but added genuine expertise and unique insight held their positions.

Notably, the vast majority of top-ranking pages now use AI assistance in some form — so the dividing line is not AI versus no AI. It is oversight versus no oversight; original substance versus rehashed summary. The update did not rewrite the fundamentals of search. It made them less forgiving of the middle ground.

Five problems that appear when you scale without a plan

If the extra volume is simply “more AI doing more of the work,” these are the predictable failure modes:

1. Quality dilution. Same human oversight spread across five times the output means average quality drops. The unedited pieces drag down how the whole domain is judged.

2. Topic thinning and self-cannibalisation. Most niches have a finite set of genuinely useful things to say. Going from monthly to weekly often means covering marginal topics, or writing several posts that target the same intent — splitting your authority instead of building it.

3. Indexing limbo. More published does not mean more indexed. Google increasingly leaves low-value pages in “Crawled — currently not indexed.” You can publish twenty posts a month and have half sitting unindexed, which is a wasted signal, not a stored one.

4. Author and trust signals get stretched thin. Google’s E-E-A-T framework now rewards content attributed to named experts with verifiable credentials. Anonymous, high-frequency output is losing ground precisely because the volume makes credible attribution harder to sustain.

5. Reader experience erodes. Brand trust falls when people land on pages that feel empty or repetitive — and reader signals now feed directly back into rankings.

The part most people miss: citation beats cadence

Here is where the conversation has genuinely changed, and where the opportunity sits.

Search is no longer only about blue links. AI Overviews, AI Mode, ChatGPT, Perplexity and Gemini increasingly answer the question on the results page itself. On many queries the click never happens. That sounds like bad news — and for high-volume generic publishers, it is. But it reframes what content is for.

In an answer-first environment, being the source that gets cited is worth more than being one of many pages that rank. The data backs this up: brands cited as sources within AI Overviews receive around 35 percent more organic clicks than brands that are not cited, and a single citation inside an AI Overview can generate more qualified traffic than ranking in position three on the same query through traditional results.

LLMs and AI Overviews do not cite the highest-frequency publisher. They cite content with clear expertise, original data, verifiable authorship and a structure that makes the answer easy to extract. A handful of deeply authoritative, well-structured pieces will out-earn a flood of competent ones — because being cited, not merely present, is now the prize.

The reassuring part: this is not a separate strategy from classic SEO. The same signals Google’s core update rewards — originality, expertise, authority, clean structure — are the signals AI models trust and cite. You are not building two things. You are building one thing well.

What more content can legitimately help with

To be fair to volume, scaling up is not automatically wrong. Done deliberately, more content can genuinely help with:

  • Topical breadth — covering real, distinct subtopics you previously had no time for, which strengthens topical authority.
  • Internal linking — more genuinely useful pages create more opportunities to connect related content.
  • Distribution cadence — a steadier social rhythm to surface and resurface your best work.

These are real benefits. They are also secondary to the substance of what you publish. Volume amplifies quality; it does not substitute for it.

The shift that actually matters: change the KPI

The single most useful change a brand can make when AI speeds up production is to stop measuring the wrong thing.

The durable model is straightforward: human expertise provides the substance, AI provides the efficiency. Subject-matter experts stay essential; the tools simply make them more productive. One authoritative, well-researched guide is worth more than twenty thin pieces covering adjacent ground.

So replace “articles published per month” as a success metric with measures that reflect what now drives results:

  • Content that earns citations in AI answers and third-party publications
  • Content that achieves real engagement — time on page, return visits, shares
  • Content that generates qualified traffic and enquiries, not just impressions

If a piece would not meet that bar, the extra speed AI gives you is better spent making fewer pieces genuinely excellent than making more pieces merely adequate.

Conclusion

Going from five blogs a month to five a week will not multiply your results — and if the volume comes from light-touch AI output, it risks the opposite, with the March 2026 data showing 60 to 80 percent traffic losses for exactly that pattern. The brands that win in an AI-mediated search world are not the ones publishing most. They are the ones producing content worth citing: original, expert-led, well-structured, and built to be the source an AI trusts.

Use AI to remove the grunt work, not the thinking. Publish less if you must — but make every piece earn its place.

Consilium Design helps ecommerce and consumer brands win visibility in AI-mediated search through feed-first and entity-first strategy. If you are scaling content and want to know whether it is helping or hurting, that is exactly the kind of question worth asking before you publish the next fifty posts.