Keyword research for a store differs from keyword research for a service business in one structural way: your catalogue already contains most of the terms. The work is less about discovery and more about deciding which page type should own each query.
The general method is on SEO keyword strategy. This page covers the ecommerce case.
The four query types, and who owns them
Category terms
“running shoes”, “office chairs”. The highest-volume commercial demand and almost always owned by a category page, not a product. These are where most ecommerce revenue is decided.
Attribute-qualified terms
“waterproof running shoes”, “ergonomic office chairs”. These are the interesting ones. They may justify a subcategory, a filtered page made indexable, or nothing at all. Getting this decision wrong is the main source of ecommerce index bloat. See when a keyword needs a category page.
Product and model terms
Brand names, model numbers, SKUs. Owned by product pages. Low volume individually, high intent, and collectively substantial on a large catalogue.
Informational terms
“how to choose running shoes”, “office chair vs stool”. Owned by guides, and they should route to categories rather than trying to sell directly.
Where to find terms you do not already have
- Your internal site search. The single most underused source. It is your customers describing what they want in their own words, including things you do not stock.
- Google Ads search terms, which show what actually converts rather than what has volume.
- Competitor category structures, which reveal segments you have not created.
- Customer service enquiries, particularly recurring questions.
- Review text, where customers describe products in non-manufacturer language.
Internal search deserves emphasis. A term searched repeatedly on your site with no results is demand you are actively failing to meet.
The decision that matters most
Not which terms to target, but which page type owns each. Most ecommerce cannibalisation comes from a category and a product, or a category and a filter, targeting the same query.
- Search the term and look at what ranks. If the results are category pages, a product page will not win it.
- Broad terms to categories, specific terms to products, and be strict about the boundary.
- Do not create a page for every attribute combination. Most have no independent demand.
- Record ownership explicitly. See keyword mapping.
Volume is a poor guide here
Ecommerce has a long tail that keyword tools report badly. Model numbers and specific product terms frequently show no volume and convert at high rates. A catalogue of thousands of products competes on the aggregate of many low-volume terms, not on a handful of head terms.
Frequently asked questions
Should every product target a keyword?
Every product you want ranking, yes, usually its name and model number. For a long tail of low-value SKUs, category pages carry the demand and per-product optimisation is not economic.
How do I choose between a category and a filter?
Whether the narrower term has independent search demand. If people search it, it may justify an indexable page. If they only use it to narrow a list, it is a filter.
Are keyword tools useful for ecommerce?
Partly. They are good for category terms and poor for the product long tail. Your own site search and Ads data are better sources for the tail.
What about seasonal terms?
Plan them months ahead, since content published during a season captures the following year rather than the current one. Keep the pages live year-round rather than unpublishing them.
Consulting CTA
If you are unsure which of your pages should own which commercial terms, book an SEO consultation for a catalogue mapping review.