Ecommerce Keyword Research: Finding the Right Terms for Your Products

Warning: fileperms(): stat failed for /www/.wordpress/ in /wordpress/wp-admin/includes/file.php on line 2225

Warning: fileperms(): stat failed for /www/.wordpress/ in /wordpress/wp-admin/includes/file.php on line 2225

Keyword research for an ecommerce store looks nothing like keyword research for a service business. Service businesses have a handful of pages targeting a handful of high-intent queries. Ecommerce stores have hundreds or thousands of pages, three or four distinct intent types per query, and a competing tension between what customers search and what your inventory system happens to call things.

After running keyword research for ecommerce stores across jewellery, fashion, consumables and gifting, the pattern that keeps repeating: most stores organise their catalogue around inventory logic rather than search demand. The fix is a structured keyword research process that maps every collection and top product to a primary query, validates demand, and rules out cannibalisation. This guide is how I do that.

This guide covers the three types of ecommerce keywords, the tools I actually use, a step-by-step research process, and how to map keywords to URLs without creating cannibalisation. It pairs with our collection page SEO guide, product page SEO guide, and the broader Shopify SEO guide.

The Three Types of Ecommerce Keywords

Every ecommerce keyword falls into one of three intent buckets, and each maps to a different page type.

Category / Collection Keywords (Browse Intent)

Examples: “engagement rings”, “women’s running shoes”, “Australian-made signet rings”, “linen midi dress”.

These are the highest-volume queries and they drive the majority of organic traffic on most ecommerce stores. The user is browsing, not yet attached to a specific product. They land on a collection page, scan the range, filter, and click through to whatever catches their eye.

Collection keywords are where SEO investment pays off most reliably. They have stable volume, predictable intent, and clear placement (one keyword per collection page).

Product / Brand Keywords (Purchase Intent)

Examples: “Nike Pegasus 41 review”, “Black Finch sapphire engagement ring 1.5 carat”, “Apero label nursing dress”.

Lower volume per query but very high commercial intent. The user has already done the consideration work and is searching for a specific product. They land on a product page and convert (or don’t) within seconds.

Product keywords cumulatively add up because there are so many of them. A store with 500 products has thousands of potential long-tail product queries. The investment is in writing unique, distinctive product descriptions rather than spending hours optimising any single page.

Informational Keywords (Consideration Intent)

Examples: “how to choose an engagement ring”, “best running shoes for flat feet”, “linen vs cotton for summer dresses”.

The user is in the research phase. They don’t want to be sold to. They want to understand the category, learn the trade-offs, and figure out what to look for. They land on blog content, read, and (if you’ve done it well) leave with your brand established as a credible source.

Informational content feeds the funnel above purchase intent. It is also where AI search engines pull source material for AI Overviews and agentic shopping experiences. Increasingly, this is where brand visibility starts.

The Tools That Matter

The tools I actually use for ecommerce keyword research, in order of how often:

• SE Ranking for keyword research with Australian-specific volume data and competitor tracking.
• Google Search Console for query data the store is already getting impressions for. Free, accurate, and shows what is actually working vs what you expected to work.
• Google Trends for seasonality patterns and emerging queries that keyword tools haven’t picked up yet.
• Google Keyword Planner for raw search volume sanity-checks, especially for new categories.
• Ahrefs or Semrush if budget allows, for competitor gap analysis and link intelligence.
• SE Ranking’s Keyword Gap tool to identify keywords competitors rank for that you don’t.

For most ecommerce stores starting out, SE Ranking plus Search Console covers 90% of the workflow. Add Ahrefs or Semrush when you outgrow that.

Step 1: Audit Your Existing Inventory

Before researching new keywords, understand what you have. Export your full product catalogue (Shopify admin → Products → Export). For each product, you need: title, type, vendor, tags, current collection assignments, and current URL handle.

The audit answers three questions:

• How many products do you have in each broad category?
• Which collections already exist and how are they organised?
• Where do products overlap between collections, and where are there gaps?

This becomes the input for everything that follows. You cannot research keywords sensibly until you know what you actually sell.

Step 2: Find Collection-Level Keywords

For each existing or planned collection, identify the primary query a customer would use to find it. Then validate with data.

Process per collection:

• Draft the primary query. Start with the obvious: “sapphire engagement rings”, “linen midi dresses”, “Australian made gold signet rings”.
• Pull search volume. Check in SE Ranking or Keyword Planner. Look at exact match, broad match, and related queries.
• Check competitor rankings. Run the query in Google AU. See who ranks. Are they direct competitors or aggregators? What does the search experience look like?
• Look at related queries. SE Ranking’s keyword suggestions will surface variations you missed: location modifiers (“sapphire engagement rings Australia”), material variations (“blue sapphire engagement ring”), size or carat modifiers.
• Pick the strongest primary and 2-4 secondary keywords. The primary goes in the H1 and meta title. The secondaries get woven into the intro copy and long-form content.

If a collection has no meaningful search demand at any reasonable primary query, either reposition it (different angle, different name) or accept that it should be a tag rather than a standalone indexable collection.

Step 3: Find Product-Level Keywords

Product keyword research is more scalable and less manual than collection research because the patterns are predictable.

For each product, ensure the title includes:

• The descriptive product name (what customers would call it)
• Brand, if it matters in your category
• Material, size or other distinguishing attribute
• Model number or SKU only if it is a known search query in your category

Title structures that work by category:

• Clothing / apparel: Brand + Gender + Product type + Colour + Size + Material
• Jewellery: Stone/Material + Style + Setting + Brand
• Electronics: Brand + Model + Storage/Capacity + Colour
• Consumables: Brand + Product type + Weight/Count + Flavour or variant

This same logic flows into Google Shopping feed titles. We cover the full feed optimisation pattern in our product feed optimisation guide.

Step 4: Find Informational Keywords That Feed the Funnel

Informational keywords are how you build topical authority and capture customers earlier in their journey. Find them by:

• People Also Ask on existing high-intent SERPs. Search your primary collection keyword, scroll to PAA, expand each. Every PAA question is a potential blog post.
• Related searches at the bottom of Google’s SERP.
• Reddit and forums for genuine questions customers ask before they buy.
• Customer service emails and live chat transcripts for the questions your team answers repeatedly.
• SE Ranking’s question keyword filter for queries containing how/what/why/best.

The best informational keywords are ones where you can write something genuinely useful and link naturally to a relevant collection or product. “How to choose an engagement ring” → links to your engagement ring collection. “How to clean a sapphire” → links to your sapphire collection and your ring care page.

Step 5: Map Keywords to URLs

The output of the research process is a keyword-to-URL map. One row per keyword, with: query, search volume, intent (browse / purchase / informational), primary target URL, page type (collection / product / blog), and current ranking position if any.

Most ecommerce stores I work with discover during this exercise that:

• Several keywords have no assigned URL (gap content opportunities)
• Several URLs are competing for the same keyword (cannibalisation)
• Several existing URLs have no strong target keyword (rewrite or delete)

The map becomes the working document for content strategy. New blog posts, new collections, page rewrites all reference it.

Avoiding Cannibalisation

Cannibalisation is when two URLs on your store compete for the same keyword. Google can’t decide which to rank, so usually neither ranks well.

Common ecommerce cannibalisation patterns:

• A collection page and a blog post targeting the same broad term (e.g. “engagement rings” collection + “engagement rings guide” blog post). Differentiate by intent: collection page targets the buy query, blog post targets the informational variant (“how to choose an engagement ring”).
• Two collection pages targeting the same query (e.g. “engagement rings” and “rings for engagement”). Merge into one.
• A tag-based collection and a manual collection covering the same products. Pick one as the canonical, redirect the other.
• A product page ranking for a query the collection page should own. This is less of a problem if both are on the first page; otherwise consider whether the product is being mistaken for the broader category.

SE Ranking’s keyword cannibalisation report or a quick spreadsheet of GSC query data per URL surfaces these quickly.

Validating Search Demand vs Real Buyer Intent

Search volume is necessary but not sufficient. A query with 1,000 monthly searches in Australia could be:

• Real buyers ready to purchase (gold standard)
• People doing research with no purchase intent yet (informational)
• People looking for free options or how-to content (poor commercial fit)
• Industry insiders, competitors, journalists (not your customer)

To validate intent, run the query in Google AU and look at:

• SERP composition. Is it dominated by product listings, blog posts, news, or YouTube videos? The SERP shape signals what Google thinks the query intent is.
• Existing top results. Are they direct competitors selling the product, or aggregator sites, or content publishers?
• SERP features. Shopping carousel, AI Overview, People Also Ask, knowledge panel. Each signals different intent.
• Featured snippets. If a featured snippet exists, the query has informational intent. Adjust your content strategy accordingly.

If the SERP is dominated by content publishers and the top product listing is at position 8, that query is informational regardless of search volume. Don’t try to land a collection page there.

Ecommerce Keyword Research Mistakes

The patterns I see most often:

• Optimising for volume without checking intent. Chasing a high-volume query that turns out to be informational or commercial-fit-poor.
• Ignoring local modifiers. Australian customers add “Australia” or city names more than international SEO advice usually accounts for.
• One keyword per page absolutism. Modern ranking is semantic. A well-written collection page ranks for the primary keyword AND dozens of secondary variations naturally.
• Skipping the inventory audit. Researching keywords for collections you can’t actually populate with enough product variety.
• Cannibalising your own pages. Building a blog post that competes with the collection page targeting the same intent.
• Ignoring Search Console. The keywords you already get impressions for (but no clicks) are the easiest near-term wins.

Seed Keyword Brainstorming Before You Open a Tool

The tool you pick is only as good as the seeds you feed it. If you start a SE Ranking search with “jewellery”, you get a generic list. If you start with “art deco engagement ring”, you get a usable shortlist.

My seed list for a new client comes from four places. The product catalogue itself, including any internal product type or tag taxonomy. The terms the client uses verbally when they describe what they sell, which are almost always more specific and more buyer-led than the words on their site. The Shopify search log, if it has been running long enough, which shows the exact phrasing customers type into the site search bar. And a quick read of customer reviews, which surface descriptive words customers use that the brand often does not.

For Apero Label that exercise produced seeds like “button down nursing dress”, “discreet breastfeeding top” and “maternity work dress”, which were all closer to actual buyer language than the on-site category names. Feed those into your tool of choice and the keyword list that comes back is already filtered to intent.

Mining Competitor Keyword Gaps

For a new Shopify store with a thin keyword list, competitor gap analysis is usually the fastest way to a useful target list.

The workflow I run for Rank Haus clients is the same every time. Pull three to five direct competitors, not aspirational ones. For Black Finch Jewellery that means other independent Australian jewellers in the same price band, not Tiffany. For Apero Label it is other Australian maternity and nursing labels, not ASOS. Drop each domain into SE Ranking’s Competitive Research, export the organic keywords they rank for in the top 20, then filter out branded terms and anything with zero search volume in Australia.

The interesting list is the intersection: keywords two or more competitors rank for that your site does not. That intersection is your demand-validated, intent-validated shortlist. Someone has already proven a Shopify store can rank for these terms, which removes most of the guesswork about whether the keyword is realistic.

Then I split that list by intent. Category-level gaps become new collection briefs. Product-level gaps go on the product optimisation list. Informational gaps go to the blog calendar. For Fanzy Pantz this exercise surfaced about 40 collection-level terms competitors were ranking for that the site had no matching page for, which then drove the next two quarters of collection work.

Search Volume, Keyword Difficulty and How I Weight Them

Search volume in Australia is small. A collection keyword doing 200 searches a month here is solid. A product-level keyword doing 50 is worth a page. I will not chase a keyword with no Australian volume just because the global number is high, because the buyer pool is wrong and the SERP is usually US-dominated. SE Ranking and Ahrefs both let you filter to AU volume specifically. Use that filter.

Keyword difficulty matters less than the competitor mix in the actual SERP. A KD of 45 looks intimidating until you open the first page and find five marketplace listings, two news articles and one weak competitor. For a real independent Shopify store, that SERP is winnable. A KD of 25 with five established niche competitors above the fold is harder. I open every shortlisted keyword in an incognito Google AU search before committing to it. Tools approximate. The SERP is the truth.

The weighting I apply: AU volume sets whether the keyword is worth pursuing at all, SERP composition sets whether ranking is realistic in the next six months, and commercial intent sets the priority order. Difficulty scores are a tiebreaker, not a gate.

Tracking and Refining Keyword Performance After Launch

Keyword research is not a one-off exercise. The list you build in month one will be wrong in places, and you only find out which places by watching the data.

For every client I set up the same tracking stack. SE Ranking holds the daily rank tracking for the priority keyword list, segmented by collection, product and informational. Google Search Console gives the impression and click data on every keyword the site actually shows for, including the long tail the rank tracker does not know about. GA4 ties the landing page sessions back to add-to-cart and purchase events for an organic-only audience.

I review the three together monthly. The pattern I look for: keywords picking up impressions but stuck on page two, which usually need an on-page fix or one internal link. Keywords ranking but not converting, which usually means the page intent is off, not the keyword. Keywords I targeted that never picked up impressions at all, which means the topic was wrong and I drop it.

For Black Finch this loop has replaced about 15 percent of the original keyword list over the last six months with terms that were either bringing in unexpected impressions or converting better than the original picks.

Using Google Search Console as a Keyword Research Tool, Not a Reporting Tool

For any site that has been live more than three months, Google Search Console should be the first place you look for keyword opportunities, not the last.

The trick is the filter. Open the Search Results report, set the date range to the last 16 months, then filter the Queries tab to impressions greater than 50 and position between 8 and 30. That single filter is the highest-leverage keyword list you will build all year. Every keyword in it is one Google already thinks your site is relevant for, with proven Australian search demand, where you are close enough to page one that a focused on-page pass can move the needle inside a month.

Then run a second pass with the same date range but filter to queries containing question words: how, what, why, can, does, is. That second list is your blog content brief. Google is telling you which informational queries it is already serving your site for, which is a much better signal than any keyword tool’s question generator.

For Fanzy Pantz this exercise produced 90 quick-win keywords on the first pull, which fed a three-month optimisation sprint that moved 34 of them onto page one without a single new page being built.

Keyword Research for AI Search and Agentic Shopping

ChatGPT shopping, Perplexity, Google’s AI Overviews and Gemini are now meaningful sources of product discovery for Australian shoppers, and the keywords that trigger them are not the same as the keywords that trigger a classic blue-link SERP.

Two patterns are showing up in the AI Overview and prompt data I track in SE Ranking for Rank Haus clients. First, AI engines pull from longer, more conversational queries: “best lab grown diamond engagement ring under 5000 in Australia” rather than “engagement rings”. Comparison and constraint-heavy phrasing wins. Second, the engines lean heavily on third-party content (review sites, guides, Reddit threads) when constructing a recommendation, so your keyword strategy now has to include the keywords that win those third-party listicles, not just the ones that win your own pages.

The practical change to my workflow: alongside the standard keyword list, I now build a prompt list. For Black Finch that is roughly 30 buyer prompts a shopper might type into ChatGPT or Perplexity when looking for a piece in their range. SE Ranking’s AI prompt tracking shows whether the site is being cited in the responses, and which competitor sites or publications are. The optimisation work that follows is partly on-page, partly digital PR, but the keyword research is where it starts.

Australian-Specific Data Sources and SERP Quirks

Most of the guides on this topic are written for a US audience, and the keyword volumes and tools they reference do not translate cleanly. A few Australia-specific notes that change how I do this work for Shopify clients here.

Use AU search volume, not global. SE Ranking, Ahrefs and Semrush all let you set the database to Australia. The keyword “engagement rings” does roughly 60,000 monthly searches globally but around 22,000 in Australia, and the SERP is entirely different. Make decisions on the AU number.

Set incognito Google searches to google.com.au with location set to your client’s primary city. Default Google results are personalised, and a SERP that looks winnable from your laptop is often dominated by US marketplaces when checked from a clean Sydney session. The free SerpWorx or Ahrefs SERP overview both let you set country and city without opening incognito tabs all day.

Watch for the marketplace mix. Australian product SERPs are heavily weighted to The Iconic, Catch, Amazon AU, eBay and Temple and Webster across most consumer categories. If five of the top ten results are marketplaces, an independent Shopify store can usually take a top-five spot with focused work. If the top ten is full of established independents, the timeline is longer. The Australian marketplace lean is, on balance, a tailwind for independent stores doing this work properly.

Want Help with Keyword Research?

If you run a Shopify or other ecommerce store and want a structured keyword-to-URL map for your catalogue, we offer a paid keyword research engagement that produces a working spreadsheet covering collection, product and informational queries. Get in touch if it would be useful.

Get in touch about ecommerce keyword research →

Frequently Asked Questions

What is ecommerce keyword research?

Ecommerce keyword research is the process of identifying the search queries customers use to find products in your category, mapping each query to the right page type (collection, product, or blog), and validating that the demand and intent justify investment. It is different from service-business keyword research because of the scale, the three distinct intent types, and the constant tension between inventory logic and search demand.

What tools do I need for ecommerce keyword research?

SE Ranking plus Google Search Console covers 90% of the workflow for most stores. Add Ahrefs or Semrush for competitor gap analysis when you outgrow that. Google Trends is useful for seasonality. Google Keyword Planner is fine as a free sanity check on raw volume. Avoid relying on any single tool — cross-reference at least two for any major decision.

How do I find keywords for product pages?

Product keywords follow predictable patterns by category (e.g. brand + style + material for jewellery; brand + gender + product type + colour for apparel). Write product titles that include the descriptive name, brand if relevant, and the distinguishing attribute. Long-tail product queries cumulatively add up across thousands of pages, so invest in writing unique titles and descriptions rather than optimising any single page heavily.

How many keywords should I target per page?

One primary keyword per page plus 2-4 secondary variations. Modern semantic ranking means a well-written page that targets one primary keyword usually ranks for dozens of related variations naturally. The “one keyword per page” rule from older SEO advice still holds for clarity of intent.

What is keyword cannibalisation in ecommerce SEO?

Cannibalisation is when two URLs on your store compete for the same keyword. Google can’t decide which to rank, so usually neither ranks well. Common patterns: collection page and blog post targeting the same broad term, two collections covering the same products with different names, tag-based and manual collections overlapping. Fix by merging duplicates, differentiating by intent (collection = buy intent, blog = informational), or 301 redirecting the weaker URL into the stronger.

How do I use Google Search Console for keyword research?

Search Console shows the actual queries your store already gets impressions for. The easiest near-term wins are queries where you have impressions but few clicks, often because you’re ranking on page 2 or 3. Pull the queries report, filter for queries with 50+ impressions and average position 11-30, and use those as the starting point for content improvements on the relevant URL.

Should I do keyword research for blog posts or just product pages?

Both. Blog content captures informational intent (the consideration stage before purchase), feeds the funnel above your collection and product pages, and is increasingly the source material AI search engines pull from for AI Overviews. A balanced ecommerce SEO strategy invests in collection, product and blog content in proportion to the search demand in each category.

How often should I redo ecommerce keyword research?

A full keyword-to-URL map review once a year, with quarterly checks on Search Console data to catch emerging queries and seasonality shifts. For categories with strong seasonal patterns (gifting, fashion, occasionwear), monthly checks during peak periods are worth the effort.