Two-thirds of Google searches now end with nobody clicking anything — not on your site, not on a competitor's. Nothing.

(That number's from Rand Fishkin's analysis of clickstream. Worth a read.)

Before you panic: most of those searches were never going to buy a brake pad from anyone.

They're mostly people looking up the weather or how to cook an egg, and AI Overviews (and tools like ChatGPT) just answer them directly instead of sending them off to read a 3,000-word blog post about someone's grandma's egg recipes.

Luckily, no amount of AI magic can conjure up, say, a cylinder head gasket for someone looking to buy one.

So, purchase-intent searches are still safe, but stores that used to get traffic on from blog posts at an earlier stage of the buyer journey, now have a smaller pool of users to woo.

So, what does this mean for you?

At a strategy level, you can go one of two ways.

Either, A:

You focus on carrying inventory that nobody else does.

A part for a vehicle that stopped being made years ago. A fitment so specific the big chains don't bother stocking it. Easier said than done, I know.

But relevance has always been – and will continue to be – the most important ranking factor. If you're the only store that carries a part, you don't need to out-market anyone. You're the #1 answer by default.

Of course, most stores can't realistically cash in on this. The demand for Karmann Ghia carburettors and R80 GS exhausts isn't yuge.

If you can't lean that way, your alternative is:

Focus on building your store brand – but not for the reasons you think.

Every store owner rolls their eyes at this, but I'm not talking about "brand" in the sense of splashy superbowl commercials.

I'm talking about what people say about your store, or more precisely, how much your store is talked about on the internet.

Linked and no-link mentions of your brand have the strongest correlation with how often your brand gets recommended by AI tools. Your SEO campaign can no longer just write a few blog posts and meta titles and call it a day.

But that's a more long term fix (more on how to do this in a future issue), so what can you do today?

Well, firstly, as Lao Tzu said, the best time to start building your brand was ten years ago, and the second-best time is now. If you're already investing in SEO, don't continue to ignore this aspect of it.

Secondly, smaller traffic numbers means each user matters more.

A page that's "good enough" — a photo, a description, a buy button — just doesn't cut it anymore.

So: pull up your three best-selling product pages. If a shopper had yours and a competitor's open in two tabs, would yours actually help them choose?

Specs that matter, fitment that's specific, what makes this one different from the other one that also "fits" – that's the gap on most aftermarket product pages.

And it's where today's three quick wins live.

Check Merchant Center. Since April, Google's been flagging product images that are too small — under 500 x 500 pixels — in the "Needs attention" tab. Here's Google's fix-it guide. Right now it's just a warning. From January 2027, it could affect whether those products show up at all. If your images come from the manufacturer, which most do, there's a decent chance some of yours are already flagged.

Translate your filters. Baymard's research found that 62% of ecommerce sites use filter labels that make sense to the site operator and nobody else. "Bolt pattern." "Offset." "Drive type." A shopper hits one of these, doesn't know what it means, and goes to Google to find out. Maybe they come back. Maybe they land on a competitor's glossary page instead, and since they're already there... they just buy. A one-line tooltip next to your trickiest filter terms can improve your sales.

Remember: machines read this too. Product pages now need to work for AI systems too, not just humans and classic search. If your fitment info only lives in a dropdown, a javascript accordion, or a PDF spec sheet, an AI can't read it. Describe what it fits, what it's made of, and how it compares in plain text on the page. One job, done once — fixes it for humans, machines, and Google at the same time. Check out tools like Hypotenuse and Describely to scale this across your catalogue.