What's left in SEO?

My best SEO post died last month. Sigh. For years, it ranked for “Growth mindset interview questions” and brought a steady stream of subscribers. Now, it’s a trickle. I had a theory… 

I typed that query into ChatGPT, and it spat out a brilliant answer – much better than my blog post – in three seconds.

No wonder everyone's using AI to automate the hard graft of SEO. But does that actually work?

tl;dr – it doesn't.

To test this, Lily Ray tracked 220+ websites who use AI content tools (sites that were cited in their case studies). She monitored their rankings and traffic using third-party tools like Ah​refs.

Her results: Initial traffic spike, but traffic fell below baseline within 6 months. Over half the sites lost 30%+ of their traffic. Google constantly updates their algorithm to "reduce unhelpful unoriginal content." (And keep you from leaving their site.)

So it’s not just me.

What to do instead?

First, don’t waste money making AI content farms – searches for generic information (e.g. “growth mindset interview questions”) aren’t coming back – even if they're written by a human. What’s left?

Think about your own search behavior: When do you scroll past an AI overview and click a link? Videos? Images? A specific tool, vendor, product, dataset, or a respected person's take? You want a specific thing, not a description of that thing.

Your new content strategy starts with a question: What specific things do you have that people search for? A book, a product, a newsletter? Proprietary benchmarks or original research? Video content that's worth watching? (You could read a summary of a MrBeast video, but it's not the same.)

If the answer is "nothing yet," that's a content strategy problem before it's an SEO/AEO problem.

You can read Lily’s full study, including 8 recurring content patterns that are risky for SEO.

I hope this helps!