Product and category keyword research

A store selling trail running shoes has its collections organised the way the stock arrives: by brand, then by the supplier's range names. The people who buy those shoes search for "waterproof trail running shoes" and "trail shoes for wide feet", and neither phrase appears in a collection title. Google can only match a search to a page that uses the searcher's language, so the keyword list has to start with how buyers describe what they want. You map it back to the catalogue afterwards.

Buyers describe the thing, your catalogue describes the stock

Google ranks a page for a search when the page's content matches what the searcher meant. It handles synonyms well, but it can't match "wide fit trail shoes" to a collection called "Summit Series" with no mention of fit on the page. Your catalogue is organised for buying stock and managing inventory; the keyword list has to be organised for the person typing.

Store keywords split into two lists that go to different pages; keep them apart from the start.

Category keywords name a type of thing, often with an attribute: "women's trail running shoes", "waterproof hiking boots", "merino running socks". The searcher wants to see a range, so the page that wins is a collection.

Product keywords name a specific item: a brand plus a model, sometimes with a size, colour or year. The searcher has chosen and wants that item, so the page that wins is the product page. Which page catches which search covers why mixing the two costs you both.

A third group, the questions people ask before they choose, belongs to guides. Note them as you go for the content chapter, but this lesson is about the two lists that sell.

Harvest the words buyers already use

Seeds come from buyers, not from a keyword database. A tool can only expand the language you give it, so start where that language is already written down.

1. Your store's on-site search log, if your platform or an app records it. Every phrase typed into your search box is a buyer saying, in their words, what they wanted, and the searches that returned nothing are the most useful lines.
2. Customer questions and reviews. "Does this fit a wide foot" is "wide fit trail shoes" waiting to be written down, and reviews name the use a product was bought for, which is often a category your store has never named.
3. Google Search Console, if the store has been live for a while. Its Performance report lists the queries your pages already appear for, including category phrasings you never targeted.
4. Google's own search box. Type a category seed and read the autocomplete suggestions, which Google's explanation of the feature says are drawn from real searches; then run the search and read the People Also Ask box.
5. Competitors' collection menus and the category trees of the large marketplaces: each is a list of category searches someone else decided were worth a page.

Write each seed in plain words. "Trail running shoes", not the supplier's range name.

Expand each seed with the modifiers shoppers add

A seed is a topic; a keyword is a phrasing. "Trail running shoes" becomes a different search, and a different shopper, with every word added to it. For stores the modifiers that matter fall into a few families.

Audience words such as men's, women's and kids'. Use words such as for racing, for muddy trails and for wide feet. Attribute words such as waterproof, lightweight and vegan. Compatibility phrasings such as "fits" or "compatible with" followed by another product. Price and availability words such as cheap, sale, in stock, and a price ceiling after "under". And brand or model names, alone or combined with any of the above.

A suggestion tool returns these expansions at scale, including long-tail keywords (longer, more specific phrases that each get fewer searches but add up, and are usually easier to win). Expect a long, messy list; the cut comes next.

Score the list, then sort it into product and category

Every candidate gets three readings. Volume is modelled monthly demand: right about scale, wrong about precision, and different in every market. Difficulty estimates how strong the current results are, and only means something compared across your own candidates. Intent, shown as a letter on the free Keyword checker (I to learn, C to compare, T to buy, N to find a site), says what the searcher wants; a store is looking for the C and T phrasings, and the I phrasings go to guides.

Cost per click is the reading people skip and stores should read first. Advertisers pay for a click when it leads to sales, so a category phrase with a healthy cost per click is one the market has already priced as a buyer.

Now sort the survivors: category phrasings with C or T intent to the collection list, brand-plus-model phrasings to the product list, questions to the content list. Keep the readings against each line, research each market separately, and treat the whole thing as a draft you reorder after the next two lessons.

What to take away

• Google matches searches to pages that use the searcher's words, so a store's keyword list starts with buyer language and is mapped back to the catalogue afterwards.
• Category keywords name a type of thing and belong to collections; product keywords name a specific item and belong to product pages; keep the lists separate.
• The best seeds come from your on-site search log, customer questions, Search Console and Google's own suggestions, not from a tool.
• Read volume, difficulty, intent and cost per click together, and treat cost per click as the market's opinion on whether that searcher buys.

Next

Now sort that list by what each searcher is ready to do in Buying intent for stores.