Should You Create Pages Just for AI Mentions? My Honest Take (And Why I’m Hesitant)

AI search rewards clarity, not content bloat. But that doesn’t mean the answer is creating a page for every possible question. Before you go down that path, here’s a real-world look at what happens when content production goes too far and how I’m currently helping fix it.

There’s a new wave of advice floating around SEO right now. Create highly targeted pages that answer very specific questions so AI tools like ChatGPT will cite or recommend your brand.

On the surface, it sounds smart.
Be the clearest answer, get cited, win visibility.

But the more I look at it, the more conflicted I feel.

Because while the idea aligns with what we’re seeing in AI search behavior, the execution could quietly create a mess for your website if you’re not careful.

And honestly, part of that hesitation is coming from what I’m actively working through with a client right now.

Let’s break this down.

What the Data Actually Says

A recent study covered in Search Engine Journal analyzed over 800,000 query and page relationships.

The big takeaway is simple.

Pages that get cited in ChatGPT are not the longest or most comprehensive.

They are the ones that:

• Match the query clearly in headings
• Answer the question directly
• Stay focused instead of trying to cover everything

That part makes sense.

We’ve been moving toward intent alignment for years. This is just the next evolution.

The LinkedIn Advice That Made Me Pause

Then I saw this post from Edward Sturm:

The idea is to create pages specifically designed to get picked up and recommended by AI.

Very targeted. Very specific. Very direct answers.

And I get it.

If AI is pulling from the clearest answer, then building content designed for that seems logical.

But here’s where I start to hesitate, because I’ve seen what happens when content strategy turns into content volume.

What I’m Seeing in the Real World Right Now

I’m currently working with a business whose organic performance has been on a steady decline.

Not because they weren’t creating content, but because they were creating a lot of it.

They have hundreds of blog posts. On the surface, it looks like a strong content library. But when you dig in, you start to see the pattern.

Same topics. Slightly different angles. Rewritten versions of essentially the same idea.

Not duplicate content in the technical sense. But functionally overlapping.

And most of it is not performing.

So what are we doing?

We’re not adding more. We’re actively reducing.

We’re auditing, consolidating, and rebuilding.
We’re identifying where multiple weaker posts should become one stronger, more intentional page.
We’re focusing on clarity, structure, and actual usefulness.

And early on in this process, one thing is already very clear.

More pages was not the answer.

Where This Can Go Wrong Fast

That’s why this new wave of “create pages for AI” advice makes me pause.

Because if we take that idea too far, we end up back in a very familiar place.

Mass page creation.

Except instead of:
“Let’s create pages for every keyword variation”

It becomes:
“Let’s create pages for every possible question AI might surface”

That creates real risks:

• Thin or overlapping content
• Keyword cannibalization
• Low engagement pages
• A bloated site structure that becomes harder to manage

And yes, potentially declining performance in traditional search if enough of those pages underperform.

I’m literally in the middle of cleaning up the aftermath of a strategy that looked a lot like this.

So it’s hard for me to ignore that.

Have I Tested This Strategy?

No.

And I want to be very clear about that.

I have not tested creating standalone pages purely for AI mentions.

Not because I don’t think it could work.

But because I’m not willing to introduce risk to a site without a clear path to measurable outcomes like leads or revenue.

Especially when I’m actively working on a site where too many low-performing pages are part of the problem we’re trying to fix.

The No-Index Idea

One thought I had was this.

What if you created these pages, but set them to noindex for Google?

In theory:

• You could test visibility in AI systems
• Without impacting your indexed site quality

But even then, I still have questions.

• Are these pages actually driving business outcomes?
• Are they worth the time to create and maintain?
• How do you measure success reliably?

If the answer is just “visibility,” that’s not enough for me.

The Real Question That Matters

At the end of the day, this is the filter I keep coming back to:

Is this strategy helping drive:

• Qualified traffic
• Leads
• Revenue

Or are we just chasing mentions?

Because those are not the same thing.

And when you’ve seen what happens when a site fills up with content that doesn’t perform, that question matters a lot more.

Where I Land Right Now

I’m not against optimizing for AI visibility.

But I don’t think the answer is spinning up a bunch of new pages.

I think the better path is:

• Strengthening existing pages
• Improving how clearly they answer specific queries
• Structuring content so it’s easy to extract and cite

In other words, make your best pages better.

Not just more.

That’s exactly the approach I’m taking with my client right now.

If You’re Considering Testing This

If you do want to experiment, I’d approach it carefully:

• Start small with a limited set of pages
• Avoid overlapping existing content
• Track performance beyond impressions or mentions
• Be ready to roll it back if it creates noise instead of results

Treat it like an experiment, not a new content strategy.

Final Thought

I don’t think the takeaway here is:
“Create more content for AI”

I think it’s:
“Create clearer content for real questions”

There’s a difference.

And getting that wrong could cost you more than it gains.

If you’re trying to grow visibility in both traditional and AI search without creating a mess, that’s exactly what I help with. Check out my SEO services or reach out.