Ecommerce SEO

Your shop's filters, like colour, size or brand, could be bringing in customers from Google. Some filters match things people actually search for, such as "black lanyards" or "Nike trainers". Those deserve their own proper page, with a clear title, heading and a short paragraph of helpful copy. Other filters, such as "in stock", "on sale" or "sort by price", are just there to help shoppers narrow things down. Nobody searches for them, so they shouldn't become pages that Google can find.

Before turning a filter or product group into its own page, check three things. First, demand: are people actually searching for it? Use a keyword tool or Google Search Console to check. Second, depth: do you have enough products to fill the page? Aim for at least six, because a page with two products looks thin. Third, intent: search the term on Google and see what comes up. If it's mostly category pages, a category page makes sense. If not, and its full of news sites, reddit threads or unrelated searches, it may not be worth creating and Google and other AIs and search engines may consider it spam.

Your product information now appears in lots of places: your website, Google Shopping, marketplaces like Amazon, and AI shopping answers. If those places disagree, for example your website shows a product in stock but your Google Shopping feed says it's sold out, Google can stop showing your products altogether. Pick a handful of your best sellers and check them across every channel: price, stock, sizes, colours and pack size. Fix anything that doesn't match, and get into the habit of updating everywhere at once.

AIs are fantastic at collating data and if they know that something is a mismatch, then it's an instant red flag.

Shoppers now search with real situations in mind, like "dry shampoo alternative for curly hair I can use after the gym". Your product pages can answer those searches with simple sections such as "Perfect for", "When to use", "Do not use with" or "What customers say". A plant shop might add "Perfect for north-facing rooms" or "Plant now for flowers in spring". These small details help customers decide, and they help Google and AI tools match your products to the very specific questions people ask.

The same product can be bought by very different people for very different reasons. A watch might be bought as a 30th birthday gift, as someone's first proper watch, or by a style-lover hunting a 70s look. Yet most product pages only speak to one of them. Think about who buys your product and why, using your reviews, customer questions and Reddit discussions. Then say clearly on the page who your product is for, and who it isn't for. AI tools love this, because it matches longer, more personal searches. Struggling to define your customer base, try a pro tip and ask AI who would or wouldn't buy this if you don't actually know.

When a product sells out, it's tempting to delete the page or hide it from your menus. Don't. That page may have built up Google rankings that disappear when it goes. If the product is coming back, keep the page live, keep it linked from its category, mark it clearly as out of stock and offer a "notify me" option or alternatives. Only remove a page if the product is gone for good, and even then, redirect it to the closest alternative or its category page.

Your product titles in Google Shopping, and increasingly in AI shopping answers, need to tell shoppers and machines exactly what the product is. A simple formula works well: brand, product name, product type, then key details like colour, size or material. For example, "Nike Air Force 1 Women's Trainers, White, Size 4.5". Put the most important words in the first 70 characters, because longer titles get cut off. A clear, detailed title helps your product match the specific searches people make. It might be a nightmare for your editorial team that would pretty editorial titles but searches are often relevant and to the point.

This helps crawlers and AI's see answers quickly without having to crawl and synthesise the whole guide.

If you go onto Amazon now, you will see this in action big time.

Many websites load important details, like prices, size guides and product information, using JavaScript after the page opens. People see them fine, but AI tools like ChatGPT often can't run JavaScript, so they never see that content. That means AI may describe your products without your size guide, or skip them because the price is missing. Ask your developer or website platform whether key content is in the page from the start. A quick check: right-click your page, choose "View page source" and search for your price.

Often things like this need you to speak to your tech team to get these things fixed.

When shoppers ask AI tools detailed questions, such as “waterproof walking boots with wide fit for winter hikes under £100”, the AI looks for products with the details to match. If your product page doesn’t mention the fit, the material or whether it’s waterproof, the AI can’t recommend you, even if your product is perfect. So list as much information as you can: sizes, dimensions, materials, weight, care instructions, compatibility, what’s in the box and who it’s for. Add answers to common customer questions too. The more complete your product information, the more questions you can answer, and the more likely AI is to mention you.

AI tools don’t only look at product pages when deciding what to recommend. They also read helpful guides that explain which products suit which people. If you sell running shoes, a guide on the best running shoes for women, or a real story of a customer training for their first marathon in your shoes, gives AI useful context it can draw on. It helps the AI understand who your products are for and when they’re the right choice. Think about the questions your customers ask before buying, then write a guide that answers each one, featuring your products naturally.

More shoppers now search with a photo instead of words. They snap a jacket they like or screenshot an outfit, and Google shows them similar products. Google Search Console now shows how your pages perform in these searches. In the Performance report, open “Search type”, choose Web, then “Multimodal”.

Use it as a scoreboard for your product photos. Pick a few products that get little photo-search traffic and upgrade their images: clear, original shots from several angles, a plain background and descriptive file names. Leave similar products unchanged. After a month, compare the two groups. If the upgraded products get more impressions and clicks, roll the same approach out across your shop.