I recently sat in on a round table with 25 SEO professionals. Most were heads of SEO or led a team of SEOs in some capacity. In other words, this was a room full of people who set content strategy for their companies.
The moderator posed a question I thought was exactly the right one to ask right now, which I interpreted as a question of quality or quantity in content production? When AI makes content production as easy and rapid as pushing a button, how should content strategies change?
What happened next told me a lot about where our industry's head is at, and what it's missing.
More content, not better content
The conversation didn't stay on quality for long. Within minutes it turned into a swap meet of AI content production tips: which tools people were using, how they'd built workflows to publish faster, how many more pages their teams could push out each month.
There was a lot of smart operational thinking in that room. But almost all of it was about making more content. Very little of it was about making better content.
Why that misses the point
The moderator's question wasn't "how do we produce more?" It was "what is content worth now that anyone can produce it?" That's a very different question.
Think about it from Google's side. Google can already generate a solid answer on almost any subject, and it does, right at the top of the results page. So what's the value to Google in crawling and indexing a page that says what its own systems could have written?
The answer is: not much. A content machine running without intention, without unique data, and without a real point of view isn't building an asset. It's producing pages Google has every reason to ignore, and increasingly, reason to classify as spam.
Google has been clear about this. It has announced four spam updates in 2026 alone: March, June, August, and a September update that began rolling out on September 24. Glenn Gabe's case studies from the August update show what got hit: scaled AI content, programmatic pages at massive scale, and thin affiliate sites with little human involvement. One site lost rankings for more than 200,000 queries.
If your content strategy is "use AI to publish more of what already exists," you're building exactly what these updates are designed to remove.
The gap: Google crawls with intent, LLMs don't (yet)
So what's the gap SEOs are missing right now? It comes down to how each system decides what to read.
Google has been at this for a long time, and it has by far the most sophisticated search algorithm out there. That algorithm is backed by the best index of pages to draw from when answering a query. Every other search engine I've researched is at least ten steps behind Google, both in the quality of its results and in how developed its algorithm is.
A big part of that advantage is how Google crawls the web: efficiently and intentionally. Google has learned that while its capacity to crawl is massive, it isn't infinite. Its own documentation says it plainly: "The web is a nearly infinite space, exceeding Google's ability to explore every publicly accessible URL." So Google is selective. It weighs a site's quality, relevance, and freshness, respects robots.txt and meta directives, and spends its crawl where the best content on a subject lives.
LLM crawlers have been way less disciplined. They tend to consume anything and everything that might be talking about a subject. Most of them ignoring the rules set by webmasters outright: in August 2025, Cloudflare delisted Perplexity as a verified bot after finding it disguised its crawler and repeatedly ignored robots.txt directives.
Without crawling parameters as mature as Google's, these systems take in the good, the bad, and the redundant in roughly equal measure. That shows up in their answers.
Where this is heading
ChatGPT and the other generative search engines may not have caught up to Google here, but they will, out of necessity. Crawling and processing the whole web costs real money and compute, and indiscriminate ingestion produces worse answers. Sooner or later, every AI search engine will have to become selective about what it crawls and what it trusts, just as Google did.
When that happens, the winners will be the sites structured to point crawlers toward their most useful content. Sites that have spent the AI era flooding their own domains with generated pages will be making that job harder, for Google and for everyone else.
What SEO leaders should do instead
The room I sat in was focused on the wrong lever. Production speed is no longer a competitive advantage, because everyone has it. Here's where I'd put the effort instead:
- Add something only you can add. Original data, first-hand experience, a real point of view. If an AI could have written your page from what's already online, Google has no reason to index it.
- Publish less, and make it count. Every thin page dilutes the signal your site sends to crawlers with limited budgets.
- Structure your site for selective crawlers. Clean architecture, strong internal linking to your best pages, and robots.txt and meta directives that keep crawlers away from low-value URLs.
- Build for where AI search is going, not where it is. Today's LLM crawlers may take everything. Tomorrow's will be picky, and they'll reward the same things Google already does.
The moderator asked the right question. Quality over quantity isn't a nostalgic preference. In the era of AI search, it's the whole strategy.
