There’s a lot of panic around AI search, and it’s deserved — but targeted. If you live on articles like «how to choose a vacuum cleaner», the AI answer really does bite into your traffic: the person read the ready answer and never came to the site. But if you sell the vacuum cleaners themselves, the story is different, and far calmer.
The reason is simple: on a product query, an AI has almost nothing to answer in your place. According to this year’s data, product queries show an AI answer in only single-digit percentages of cases — against the majority in reference and B2B topics. The store ended up in a protected niche. Not because it earned it, but because its query is, by nature, about an action rather than an answer.
Why «buy» is protected and «how to choose» is not
An informational query is closed by text: you ask, you get a paragraph, you’re satisfied. A product query can’t be closed that way. «Buy sneakers» isn’t about a fact — it’s about a choice: size, color, price, availability, delivery, whether you can return them. An AI will suggest a model, but the person still goes to place the order where there’s a storefront, a cart, and live stock. You can’t replace a purchase with an answer — it still has to happen somewhere.
This isn’t a reason to relax
The prize shifted. The fight used to be for a spot in the ten blue links. Now it’s for the storefront being the best one by the time the person reaches the choice: fast, with honest stock, reviews, decent product pages. Attention to the store hasn’t vanished — it just arrives a step later, already carrying the intent to buy.
That’s why commercial SEO — product pages, categories, filters, product markup — becomes more important than a polished blog. The blog now mostly feeds the AI; the store feeds you. It makes sense to invest where the purchase happens, not where the answer gets carried off without you.
Where the AI does reach the store after all
Honestly: not all commerce is equally sheltered. At the «just looking» stage — which brand to take, which of two is better — the AI is already involved. And if there’s nothing about your product and brand out there, you simply won’t be invited into that conversation: it recommends those there’s something to say about.
So a brand name, mentions, reviews on other platforms — that’s the same trust-building work as in broad SEO, just aimed at the moment «when I get recommended». The storefront decides things once the person is already choosing; reputation decides whether you make the shortlist before that.
What a store should do
Don’t panic, and don’t stop investing in the site after hearing that «SEO is dead». For a store it isn’t dead — it moved closer to the storefront. Put the product pages, filters, and categories in order, along with product markup, speed, and the mobile version. Work on brand reputation where the AIs take their opinions from. And stop measuring success by a top-10 position — measure it by whether the people already ready to buy can find you.
The AIs scared the whole internet equally, but hit it unevenly. The store got, arguably, the softest blow — a shame not to use that while competitors panic and rewrite their blogs.