Keyword research for ecommerce is the process of finding the searches shoppers make when they’re ready to compare or buy, then deciding which level of your store should own each one: a category page, a product page, a filtered landing page or a buying guide. The research part is easy. The deciding part is where stores win or lose, because a good keyword on the wrong page type rarely ranks.
So this guide isn’t a list of tools. It’s the set of if/then rules I use to place every commercial keyword, with an illustrative rug store as the running example. I made the store up, and I don’t quote volumes for it.
Why Is Ecommerce Keyword Research Different From Blog Research?
Blog keywords are questions. Store keywords are mostly nouns with qualifiers: “wool rugs”, “8×10 wool rug”, “washable runner rug”. The searcher already knows what they want and is narrowing down.
That changes the job. Completely. On a blog, I group questions into posts, which I walk through in my guide to keyword research for blog posts. In a store, the page types already exist, and each keyword has to slot into the hierarchy you’ve built. Get the slot right and the page ranks with modest effort. Get it wrong and you’ll fight your own site for years.
The Four Conditions That Decide Where a Keyword Goes
Every ecommerce keyword I review gets the same four checks:
- What does the keyword name? A product type (“wool rugs”) points to a category. A specific model or SKU-level phrase points to a product page.
- Which modifier does it carry? Brand, material, size, colour, “best”, “cheap” and “near me” each push the keyword somewhere different.
- What does Google already show? A grid of category pages, single product pages, or articles. The SERP overrides my guesses.
- Do you stock enough to back it up? A landing page for “jute rugs” with two products is a thin page, whatever the demand.
The first two come from the keyword itself. The last two come from outside it, and they’re the ones people skip, usually because a spreadsheet full of volumes feels like enough evidence when it’s really only half of the picture.
Category, Product or Modifier: How Do You Tell Them Apart?
Here’s the rug store’s raw list sorted into types. Labels first. Examples second.
| Keyword type | Example (illustrative) | Usual owner |
|---|---|---|
| Category (head term) | wool rugs | Category page |
| Subcategory | wool runner rugs | Subcategory page |
| Product | [brand] [model name] wool rug | Product page |
| Size or material modifier | 8×10 wool rug, jute rug | A filter, unless it passes the tests below |
| Brand modifier | [brand] rugs | Brand page, only if you stock the brand in depth |
| “Best” or comparison | best washable rugs for pets | Buying guide or blog post |
| “Near me” | rug store near me | Store location page, only with a physical shop |
Product keywords are usually the longest and the most specific. They’re also the easiest to rank, because the page that owns them is often just your product page with a clear title.
The Main Path: Map Each Keyword to the Store Hierarchy
Most keywords on a typical list are category or product terms, so this path covers the bulk of the work. I place them top-down, starting with the broadest product types and working toward individual models, because a mistake at the top of the tree spreads to every page below it.
If the keyword names a product type and the SERP shows category pages from other stores, it belongs to a category or subcategory. That’s the bulk of your money terms. My guide to category page SEO covers how to optimise those pages once they own a term.
If the keyword names a specific product or model, it goes to that product page. Put the exact name in the title and H1, and don’t build a second page for it.
If the keyword doesn’t fit any existing node, ask whether the store is missing a category. Sometimes the research shows that shoppers group products differently than your merchandiser did. That’s a structure question, and I cover how to build the tree from keyword groups in my guide to ecommerce site structure.
The default is simple. Broad and plural goes up the tree. Specific and singular goes down.
Branch 1: When a Modifier Keyword Wants Its Own Page
Modifiers cause most of the trouble. Size, material, colour and brand, in that order of pain. Your filters can already produce “8×10 wool rugs”, so the question is whether that combination deserves an indexable landing page or should stay a filter.
My rule has three tests, and a keyword needs all three:
- Real demand, checked in Search Console, Google autocomplete or Merchant Center data, not only in a third-party tool.
- At least a handful of in-stock products that match, so the page isn’t thin.
- A SERP that shows dedicated pages for that exact combination, not just the parent category.
If any test fails, the modifier stays a filter, and the parent category mentions it in copy. A store that turns every filter into an indexable page ends up with thousands of near-copies, and Google’s own guide on designing ecommerce URL structures is mostly about avoiding that crawl waste. I cover when that waste actually starts to hurt indexing in crawl budget: does it matter for your site.
Branch 2: When the Keyword Says “Best” or “Vs”
“Best washable rugs for pets” looks commercial. It is. Mostly. But when I check the SERP for queries like this, the top results are usually reviews and buying guides, not category pages.
If the SERP shows articles, write a buying guide on the blog and link it to the matching category. Don’t stuff “best” into a category title and hope. If the SERP shows a mix of stores and articles, the category can try for it, but I’d still back it with a guide. The guide answers the comparison, the category catches the click once the shopper decides, and a guide worth citing is one of the best link building tactics for an online store.
This is the branch that most differs from SaaS research. In my post on keyword research for SaaS, comparison terms are the money pages. In a store, they’re usually support pages that feed the category.
Branch 3: When the Keyword Says “Near Me”
This one’s quick.
If you have a physical shop, “rug store near me” belongs to a location page and your Google Business Profile, not a category page. The research method is different, and I cover it in my guide to keyword research for local SEO.
If you’re online only, skip these. Local results are built around places a customer can visit, and forcing “near me” into product copy reads badly.
Where Do Amazon and Google Shopping Data Fit?
Marketplace data is useful. For ideas only. Amazon’s search box suggests how shoppers phrase product searches on Amazon: sizes, materials, use cases like “rug for under bed”. I copy those phrasings into my list as candidates. I don’t treat them as proof of Google demand, and I never quote a volume for them. Amazon shoppers and Google searchers aren’t the same crowd, and a phrase that’s popular on one can be rare on the other.
Google’s own shopping data is more useful if you sell through Merchant Center. Its Analytics summary has a section on what people are searching for, with search topics and search terms you can filter by country and category. Google’s help page says the popularity figures are relative, where 100 is peak popularity and 50 is half as popular. That makes it good for spotting rising product types, not for counting searches.
Search Console helps too. Don’t skip it. For eligible stores, it adds Shopping reports alongside the Performance report. But the Queries tab is still the first place I look, because it shows which URL Google already picks for each of your commercial terms.
How Do You Stop Category and Product Pages Competing?
Pick the owner before you write. Then check that Google agrees. The usual conflict is a single product page ranking for the broad category term, or two near-identical categories splitting one keyword. I’m covering the full fix in a separate guide to keyword cannibalization.
For research, my check is quick. In Search Console, filter the Queries tab for a category term, then open the Pages tab. If more than one URL shows impressions for that term, the map isn’t clear yet. Usually the category copy and internal links need strengthening, as my category guide explains, rather than the product needing a rewrite.
Edge Cases That Change the Path
- Out-of-stock products: the keyword still belongs to the product page if the item returns. If it’s gone for good, the keyword moves up to the category. My guide on out of stock products SEO covers the page side.
- Variants: “blue 8×10 wool rug” and “grey 8×10 wool rug” usually belong on one product page with variant options, not separate pages. Separate pages for each colour create the problems in my post on duplicate content in ecommerce.
- Brands you don’t stock: skip them. Ranking a page for a brand you can’t sell sends shoppers straight back to Google.
- Seasonal terms: “outdoor rugs” may only matter for part of the year. Keep the category live all year so it doesn’t lose its history, and plan the copy refresh before the season starts.
Keyword Research for Ecommerce: The Decision Matrix
| If the keyword names… | And the SERP shows… | Then it belongs to… |
|---|---|---|
| A product type | Category pages | Category or subcategory page |
| A specific product or model | Product pages | That product page |
| A product type plus size, material or brand | Dedicated pages for that combination, and you stock enough | Indexable filtered landing page |
| A product type plus size, material or brand | Only the parent category | A filter, mentioned in category copy |
| “Best” or “vs” | Reviews and guides | Buying guide linking to the category |
| “Near me” | Map results | Location page plus Business Profile, or skip if online only |
If you only read one part of this post, read this table. Every branch above collapses into one of its rows, and when a keyword doesn’t fit any row cleanly, that’s usually a sign the store’s hierarchy needs a new node rather than a sign the keyword is bad.
When Is It Time to Get Help?
If your store has hundreds of categories, heavy faceted navigation, or keywords already split across several URLs in Search Console, the mapping gets slow by hand. That’s when I’d bring in someone who does this daily. My ecommerce SEO service handles the store side, and my keyword research service covers the research and mapping on its own.
Frequently Asked Questions
What Is the Difference Between Category Keywords and Product Keywords?
Category keywords name a type of product, usually in the plural, like “wool rugs”. Product keywords name a specific item, model or SKU. Category keywords tend to have more searches and more competition. Product keywords are narrower and often easier to rank.
Can I Use Amazon Autocomplete for Ecommerce Keyword Research?
Yes, as an idea source. It shows how shoppers phrase product searches on Amazon, which can reveal sizes, materials and use cases you missed. It doesn’t tell you how many people search Google for the same phrase, so check demand elsewhere before you build a page.
Should Every Product Filter Have Its Own SEO Page?
No. Only filter combinations with real demand, enough matching products and a SERP that shows dedicated pages deserve an indexable landing page. Everything else should stay a normal filter.
Where Should “Best” Keywords Go on an Ecommerce Site?
Usually in a buying guide or blog post, because Google tends to show reviews and guides for those searches. Link the guide to the matching category so shoppers who decide can buy straight away.
Last updated: October 2026 by Mizanur Rahman
