Your boots are waterproof. Your page just doesn’t say so.

One shopper, two ways to buy

Say a shopper wants waterproof boots under $150, in a size 10, delivered by Friday.

On their own, they’d open a few tabs, look at the photos, pick a size from the menu, hover over the little ⓘ next to “Returns,” and decide. About 90 seconds and one small sigh.

Now they ask an AI assistant instead. It doesn’t browse the way a person does. It reads each page, checks it against what the shopper asked for, and recommends the ones it can back up. So the question for a retailer isn’t “does our page look good?” It’s “can the assistant find every fact the shopper asked about?”

Press the button below to see the difference.

The design didn’t change, and it’s the same boot. Four facts moved out of a picture, a menu, a vague phrase and a pop-up, and onto the page as plain words. In this example, that was enough to get it recommended.

Four places good facts hide

Most product pages have the facts. The trouble is where they’re kept. A person finds them by looking, clicking or hovering. Many AI assistants read the page as text and don’t hover or click. Even the ones that open a real browser often skip pop-ups and menus. We see the same four problems on most pages we audit.

3. In a vague phrase. “Ships fast” sounds good, but it isn’t a date. If the shopper asked for Friday, the assistant can’t match “fast” to Friday, so it is likely to pick the store that says “Arrives Thursday.” It’s the same with “great fit” versus “runs half a size small,” or “easy returns” versus “free returns within 30 days.”

4. In a drop-down menu. Many pages show stock only after you pick a size. The assistant often doesn’t pick. It sees a menu and no stock. Write it on the page: “Size 10: in stock.”

A shopper will put up with a vague page. An AI assistant often skips it.

Here’s the full list for the boot above.

Price is almost always fine. Retailers have had to get that one right for years.

Two places count, not one. There’s what’s written on the page, and there’s the product data built into the page’s code. Shoppers never see the second part. AI assistants are built to read it, because it states the price, sizes, stock, delivery and returns in a form a machine can trust. When we audit a page, most of the gaps are usually in this hidden layer, and it is often the quickest part to fix.

How big is this, really?

Big enough that the largest names in search and AI are building shopping straight into their products.

Google explains how its AI Mode picks products. When you ask for something specific, it runs several searches at once to work out what a good product needs, then suggests products that match. It checks your product against the shopper’s list. If a fact isn’t on your page or in your product data, it can’t match it.

Can you fix this before Black Friday?

Yes, if you start now. Black Friday is November 27. You don’t need a redesign, only small template changes. Most of the gaps we find are quick wins: a delivery date instead of “ships fast,” the returns policy on the page, stock shown for each size, a summary of reviews. Fix those first and leave the bigger jobs for the new year.

Miss Black Friday, and you haven’t missed the season. The weeks before Christmas are when shoppers ask for things like “a gift for my dad that arrives before the 24th.” An AI assistant can only answer that if your page gives a last order date it can check, such as “Order by Fri, Dec 18 for Christmas delivery.” Put your holiday shipping cutoffs on the page. Then come the after-Christmas sales, when people ask for the best deal on something specific. A quick win you make in November is still working in January.

How many facts does your best seller spell out?

Open one of your best-selling product pages. Tick a box only if the fact is written on the page itself, not just in a picture, a pop-up, a tab or a menu.

We run the audit for you. In seconds.

Send us one of your product page links. We run the audit in seconds. It checks the page for 197 product details that our research shows AI assistants check, gives a score, and lists the gaps with the quick wins marked. Then, in a 30-minute session, we walk you through the results and plan the fixes with you. You don’t install anything or give us access to your systems.

What you get from Syntasa

Syntasa Product Intelligence makes your product facts complete and easy for AI assistants to read. It works in three steps:

• Audit: checks each product page against 197 product details that our research shows AI assistants check, and marks the quick wins.
• Enrich: fills the gaps from your own product, stock and delivery data.
• Serve: sends the facts to your site and its product data, and keeps them current, so AI assistants read today’s price, stock and delivery date.

Your first step: send us one of your best-selling product pages. We run the audit for you in seconds, then walk you through the results and the quick wins in a 30-minute session. After that, we can give your team access to the self-serve audit agent to check the rest of your catalog.