SEO
Build the technical, architectural, and content systems that make your store easier to discover in organic search.
Explore SEO ↗Build commercially relevant keyword clusters around products, customers, problems, attributes, and buying intent.
eCommerce keyword research is the process of finding the searches that can grow organic revenue, then turning those searches into a prioritized plan for the pages your store should build or improve. Ranking for more keywords and growing traffic are useful outcomes, but they are not the finish line. The goal is to attract search traffic that can eventually make money.
Start with a five-step process that moves from understanding the business to publishing the right pages.
Study the brand, customers, products, USPs, reviews, competitors, ranking gaps, and the different reasons someone might search for what you sell.
Turn qualitative findings into a broad keyword set, add current and competitor rankings, then group phrases that share the same search intent.
Decide which existing or new page should satisfy each keyword cluster: homepage, collection, product page, blog, comparison page, support page, or another destination.
Order the roadmap by relevance, business value, search intent, ranking feasibility, product fit, catalog depth, and strategic objectives instead of search volume alone.
Build or improve the pages, connect them with internal links, publish them, measure rankings and revenue, and use the results to decide what to build next.
Your final deliverable should not be a giant spreadsheet of keywords. It should be a guided build plan that tells your team what opportunities exist, which page should own each one, what needs to be created or improved, and what should happen first.
Most brands do not fail because there are no keywords available. They fail because the research stays too shallow, the wrong page gets assigned to the keyword, or the final page is not useful enough to deserve rankings.
Teams chase the obvious high-volume terms, even when those keywords are extremely competitive or only loosely relevant. The hidden opportunity is often deeper in the long tail: specific products, personas, problems, attributes, use cases, comparisons, and buying concerns that automated exports tend to miss.
Finding a keyword is only half the battle. A blog cannot reliably replace a collection page when shoppers want to browse products. A homepage should not be forced to target every category. And multiple pages should not compete for the same search intent.
A page is not useful because it contains the keyword. Weak content is not actionable, accurate, opinionated, relevant, or meaningfully different from what already exists. Research only creates value when the page built from it is worth ranking and worth reading.
Before opening a keyword tool, build a qualitative understanding of the business. We call this Search Opportunity Discovery: the research that tells you what the brand can credibly rank for, what customers care about, and where the most commercially useful search opportunities are likely to exist.
Start with men's running shorts. “What you sell” can reveal 5-inch shorts, 7-inch shorts, 2-in-1 shorts, and shorts with pockets. “Who you serve” can reveal shorts for marathon runners or tall men. “Problems you solve” can reveal anti-chafing running shorts. “USP and attributes” can reveal lightweight, breathable, reflective, quick-dry, or black running shorts.
The products may be the same. The search journeys are different. That is the raw material for your keyword list and, later, your site architecture.
Now turn the qualitative research into search data. The objective is breadth first: collect the different ways customers describe the product, category, problem, use case, attribute, audience, comparison, and question. Then cluster related phrases so you can analyze them without creating a separate page for every wording.
Scrape the brand website, export first-party reviews, collect competitor reviews, and pull supplemental material such as product documentation, FAQs, support content, specifications, and other language customers use to describe the product.
Search product attributes, customer phrases, USPs, pain points, People Also Ask questions, Reddit threads, comparison language, and use cases in a keyword tool. Then add the site's current rankings from Ahrefs or Semrush and Google Search Console, plus ranking keywords from relevant competitors.
Group phrases that can be satisfied by the same page. Use SERP overlap and the meaning of the query to avoid building multiple URLs that compete for the same intent.
Assign each cluster to an existing or future URL. Give each target page one primary keyword, one secondary keyword, and roughly 5-10 tertiary keywords that reinforce the same intent without creating duplicate content.
Keyword tools estimate demand. They are especially unreliable at the lower end of the volume curve, where specific commercial searches can be underreported or grouped into another phrase. A 0-MSV keyword can still be valuable when the query is highly relevant, has clear purchase intent, and describes exactly what your catalog sells.
This is where low-competition, high-conversion opportunities often hide. Use reported volume as one signal, not a permission slip.
Once you have the keyword list, decide what kind of page should rank. This is where search intent becomes important: the search result tells you what the customer expects to find after they click.
Finding a keyword without matching intent is like recruiting an army and sending the wrong unit into battle. You can have a great keyword and still lose because the page format is wrong.
A category-defining keyword is the broad commercial phrase that best explains what the business sells as a whole. For a focused store, the homepage may target something like “airsoft guns,” “roses,” or “office furniture.” More specific searches should usually belong to dedicated collection or product pages so the homepage does not compete with the rest of the site.
For many eCommerce brands, the highest-value keyword opportunity is a missing commercial landing page. Use this test before adding another collection:
A strong architecture usually moves from broad to specific: Homepage → Pillar Collection → Subcollection → More Specific Subcollection → Product. See our eCommerce site architecture guide for the hierarchy and our collection-page SEO guide for how to build the destination.
Once every cluster has a proposed destination, turn the page map into a roadmap. Search volume can help estimate opportunity, but it should not decide the order by itself. A smaller commercial keyword that your store can win may be more valuable than a giant informational term that never converts.
Start with the intersection of relevance, revenue potential, and realistic feasibility. SEO should compound from wins—not begin with the hardest keyword in the spreadsheet.
Keyword research is only useful when it changes the website. Build or improve the assigned pages, connect them to the rest of the cluster, and measure performance at the page and query level.
See our eCommerce internal linking guide for the linking system that connects these assets once they are live.
Yes—when the topic, page format, product, and conversion path are aligned. Lone Pouches is a useful example because the content strategy did not start with “nicotine blog topics.” It started with customer concerns and product differentiation.
The campaign focused on non-branded searches around cleaner, safer, and healthier product alternatives—queries that aligned with the product's positioning and with people actively comparing options.

At a $63.05 average order value, 618 checkout-completing blog sessions represent approximately $38,964.90 in attributed revenue. Against the $2,500 content investment, that is about 15.6x revenue-to-spend and an estimated 1,459% ROI using the standard formula: (revenue − cost) ÷ cost.
The lesson is the alignment behind the keywords: customer concern + product differentiation + realistic target + correct page format + useful content + conversion path. Traffic became valuable because the content solved a real search problem that naturally led back to the product.
Carbon Box Media turns keyword research into the collections, content, internal links, and implementation roadmap behind the campaign. Book a strategy call and we will review where your store is leaving search demand on the table.
Once the keyword map is approved, execution moves to the page itself. The target keyword determines which URL owns the intent; on-page optimization makes that URL the clearest, most useful destination for the search.
That usually means aligning the URL, meta title, H1, heading structure, above-the-fold copy, supporting entities, collection or product depth, FAQs, internal links, and conversion path with the page's assigned keyword cluster.
Continue with our eCommerce on-page SEO guide to see how we execute a mapped keyword on the page after research is complete.
eCommerce keyword research identifies searches that can contribute to organic revenue and maps those searches to the right homepage, collection, product, landing, comparison, support, or content page.
No. Search-volume tools are estimates and can underreport highly specific searches. Validate the query's relevance, intent, SERP, and business value before deciding whether to keep or remove it.
Create a collection when the query has commercial browsing intent, the assortment genuinely satisfies the search, there are at least 5 relevant products to make the page useful, and the intent is distinct enough to deserve its own destination.
Give each page one clear primary intent, but group closely related variations that lead to the same SERP and can be satisfied by the same page. Do not create a separate URL for every wording.
Compare the live ranking domains, page types, content quality, page-level links, topical depth, and your own product advantage. Keyword difficulty is useful context, but it is not a universal yes-or-no score.
Yes, when the link helps the reader. Supporting content should move users toward relevant commercial destinations and show search engines how informational topics connect to the products and categories the store wants to rank.
AI is useful for expanding semantic variations, personas, problems, mechanisms, use cases, and content angles. It should support—not replace—customer research, search data, product knowledge, SERP validation, and human judgment about business value.