eCommerce

Keyword research for ecommerce product and category pages

8 min readBy Jaskirat Singh Talwar

Hands sorting blank paper product tags tied with orange string on a wooden table

Keyword research for a store is the work of finding the words buyers use for what you sell and deciding which single page should answer each one. Broad product-type searches belong to category pages; searches with a brand, model, part number or exact specification belong to product pages. Start from your own data (site search, Search Console, customer language), check the live results page to confirm intent, and keep a keyword map so two pages never compete for the same search.

Key takeaways

  • Collect terms from your own sources first: on-site search, Search Console queries, customer emails and reviews, and supplier specifications. Keyword tools come second.
  • Check intent by looking at the results page itself: if Google shows category-style pages, the term needs a category page; if it shows a single product, a product page.
  • Give each search intent one page. If a category and a blog post chase the same term, merge them or retarget one.
  • Search volumes from tools are estimates. Use them to rank terms against each other, not as a forecast of traffic.
  • Search Console impressions, CTR and position were inaccurately reported from 13 May 2025 to 27 April 2026, so compare periods across that window with care.

Why is ecommerce keyword research different?

Because a store has two kinds of pages that answer different searches. A category page answers "show me the options": someone searching "women's waterproof hiking boots" wants to compare. A product page answers "I know what I want": a model name, a part number, a size and colour. The 2018 version of this post made the same point with a simple example: a person searching "shoes" may be browsing, while someone searching "women's brown boots" knows what they want. That is still the core of it. The extra work in a store is scale: hundreds of categories and thousands of products, each needing the right words.

Google's own ecommerce guidance adds a structural reason. It says Google judges a page's relative importance by how many links point to it within the site, not by its URL structure. Your keyword research therefore decides which pages your menus and category links should favour. Our ecommerce SEO guide covers what to do on those pages once the research is done; this post covers the research itself.

Where do you find the words your buyers use?

Start with sources that reflect real buyers, then widen:

  1. On-site search logs. What people type into your store's search box is a list of product words in their language, including misspellings and terms your menus do not use. If a common search returns nothing, you may need a new category or a synonym.
  2. Search Console. In the Performance report, group by query and by page to see which searches already bring impressions to each category and product. Queries that show impressions but few clicks are often the quickest wins.
  3. Customer language. Reviews, support emails, chat transcripts and returns reasons show how people describe products and problems ("fits wide feet", "for a 2019 model").
  4. Supplier and product data. Model numbers, part numbers, compatibility lists, materials and sizes. These are what product-page searches contain.
  5. The results page. Google's autocomplete suggestions, related searches and the "People also ask" box show how searches branch.
  6. Keyword Planner. Google's tool gives keyword ideas and estimates of monthly searches and cost. Google requires a Google Ads account with billing information set up to use it.
  7. Competitor and marketplace categories. How larger stores name and group the same products shows which terms buyers already recognise.

How do you decide whether a term needs a category or a product page?

Search the term and look at the first page of results. The results show what Google has learned searchers want.

What the results page shows Page type to target Example shape
Category and collection pages from several stores Category page "linen bed sheets"
One product type with an attribute, shown as filtered lists Category page, or an indexable filtered page if demand is clear "king size linen bed sheets"
Individual product pages, often the same model across stores Product page brand plus model, or a part number
Guides, reviews and articles Buying guide or article that links to the category "how to choose bed sheets"
Maps and local results A location or store page "bed linen shop near me"

If the results are mixed, the intent is mixed. Choose the page type that matches most of the first page, and link to the other.

How do you choose between terms?

Rank candidate terms on four things, roughly in this order:

  1. Fit with what you sell. A term with high volume for a product you do not stock is worth nothing.
  2. Purchase intent. Specific terms (type plus attribute, model, size) usually sit closer to a purchase than broad ones. The 2018 post put it well: would you rather have 100 searches a day from people ready to buy or 500 from people browsing?
  3. Margin and stock. Put effort behind products you can keep in stock and that make money.
  4. Competition on the results page. If the first page is entirely large brands and marketplaces, a more specific term ("handmade borosilicate glass storage jars" rather than "glass jars") is often more realistic for a smaller store.

Volume comes last, and only as a comparison: Keyword Planner's figures are estimates. Google's guidance on helpful content also warns against content made primarily to attract visits from search engines and against writing to a particular word count, so a term earns a page only when you have something useful to sell or say for it.

How do you build and use a keyword map?

A keyword map is a sheet with one row per page: the URL, its main term, two to five close variants, the intent, and notes. Rules that keep it useful:

  • One main term per page, one page per main term. Search your map before creating any new page.
  • Group close variants. Singular and plural, word order and synonyms ("sofa" and "couch") usually belong on the same page, not separate ones.
  • Decide on filters deliberately. An attribute with real search demand (a material, a size range, a use) can justify an indexable filtered page with its own heading and copy; most filter combinations should not be indexed. The ecommerce SEO guide explains how to handle the rest.
  • Write the words into the page. The category's heading, title, intro and product names should use the term buyers use, not a supplier's internal name.
  • Link by priority. Menus, category pages and buying guides should link most to the pages that matter most, because internal links signal importance to Google.

For product pages, the "keyword" is often the product's own data: brand, model, size, colour, compatibility. Getting those fields complete and consistent in the catalogue does more than writing new copy.

How do you check that it worked?

Use Search Console's Performance report, filtered to the page, and compare queries, impressions and clicks over a few months. Two cautions. First, Google logged a data anomaly: impressions, CTR and average position were not reported accurately from 13 May 2025 to 27 April 2026 (clicks were not affected), so a before-and-after that crosses that window needs a note. Second, rankings and clicks move with seasons and with competitors, so compare like periods and look at clicks and orders, not position alone. Our Search Console guide explains the report.

Repeat the research when you add a range, when on-site search shows a new pattern, and before seasonal peaks.

How can Impinge help?

We build, migrate and run online stores on Shopify, Magento, WooCommerce and BigCommerce, including the catalogue and product data that keyword research feeds. See ecommerce development to talk through your store, or SEO services for ongoing search work.

FAQ

How many keywords should each category page target?

One main term and a few close variants that mean the same thing. If two terms need different results, they need different pages.

Should I target the keywords my competitors rank for?

Only where you sell the same thing and can offer a better page. Use competitor categories to learn the words buyers use, not as a list to copy.

Are long-tail keywords worth it for a small store?

Usually yes. Specific terms have fewer searches but clearer purchase intent and often weaker competition, which makes them more realistic for a small store than broad product terms.

Do I need a paid keyword tool?

Not to start. Your on-site search, Search Console, customer language and the results page itself cover most of what a small store needs. Keyword Planner adds volume estimates once you have a Google Ads account with billing set up.

Should product descriptions repeat the keyword many times?

No. Use the buyer's words naturally in the title, heading and description, and complete the product data. Google does not reward writing to a word count or repeating terms.

How often should keyword research be redone?

Review the map when you add products or categories, before seasonal peaks, and when Search Console or on-site search shows new terms.

Sources (5 links)
Share this article