On Google UK, "wine online" gets 1,300 searches a month. "Champagne" gets 49,500 and "red wine" 18,100, and the people typing them want to buy just as much as the first group. Majestic ranks first for all three with category pages, and that is where most of its estimated 637,851 monthly organic visits come from (DataForSEO, Google UK, September 2026). The SEO of an online shop is decided almost entirely by this difference: demand does not sit in keywords with "online" in them, it sits in categories.
Ecommerce SEO is the work that gets an online shop's category pages and product pages onto the first page for the searches of people about to buy, and that stops the catalogue itself (filters, variants, internal search) from getting in Google's way. Compared with SEO for a brochure site, three things change: the pages that matter number in the hundreds and look alike, the structure is partly decided by the platform (Shopify, WooCommerce, PrestaShop), and the result is measured in orders, not visits.
Below is the method we apply to the online shops we look after, with figures from UK leaders (Majestic, The Body Shop, Laithwaites, Naked Wines) and from two Shopify stores we manage, Scovaventi and Bosatta. It is a working guide; the service itself is described on our ecommerce SEO page.
How online shoppers search, according to the SERPs
We read the wine and natural skincare SERPs in September 2026 (desktop, Google UK). For "wine online" the organic results are majestic.co.uk, winedirect.co.uk, stellarcellar.co.uk, oakhamwinesonline.co.uk, vinissimus.co.uk, nakedwines.co.uk (twice), reservewines.co.uk and a second Oakham Wines page. All wine shops. No generalist marketplace, no newspaper.
For "natural skincare products" the picture is similar: six brand sites (Neal's Yard Remedies, Origins, Dr. Hauschka, Green People, The Edinburgh Natural Skincare Company, Botanicals), one multi-brand shop (mypure.co.uk), a travel blog and a Marie Claire list. No aggregators, and Amazon does not appear in the organic results.
Neither SERP has a local pack or text ads, and both open with an AI Overview. The skincare SERP also has two "Popular products" blocks, 10 product cards each, fed by Google Merchant Center; the wine SERP has none. The sellers in those blocks include Amazon.co.uk, Boots.com, Holland & Barrett, PriceRunner, Neal's Yard Remedies, Green People and small shops such as The Natural Beauty Pot and The Herbary. For a niche shop that block is worth as much as an organic position, and getting into it takes a well-built feed, not size: we cover this in the companion article on Google Ads for ecommerce.
The figure that matters most is volume. "Wine delivery" gets 2,900 searches a month, "wine online" 1,300, "buy wine online" 1,000, "online wine shop" 480. Category keywords without "online" are in another league: "champagne" 49,500, "prosecco" 49,500, "red wine" 18,100, "crémant" 14,800, "barolo" 8,100. Majestic ranks in the top three for each of them, with category pages. If you only optimise for keywords with "online" or "buy" in them, you are looking at a small fraction of demand.
The same applies across borders: a version of the catalogue with a local domain, local currency and local delivery terms picks up demand that large foreign retailers leave uncovered.
Recommended reading: how to read the search intent behind a keyword.
Which pages really bring in the traffic
Categories. You can see it by taking each leader's hundred highest-traffic keywords and looking at where they land.
| Domain | Organic keywords | Estimated organic visits per month | Keywords in the top 10 | Where the 100 highest-traffic keywords land |
|---|---|---|---|---|
| majestic.co.uk | 21,593 | 637,851 | 7,112 | 61 categories, 15 homepage, 15 guides and blog, 6 products, 3 store pages |
| thebodyshop.com | 6,225 | 188,265 | 2,418 | 56 categories (/collections/), 16 products, 12 offers pages, 8 store locator, 7 homepage, 1 blog |
| laithwaites.co.uk | 8,406 | 163,754 | 3,366 | n/a |
| nakedwines.co.uk | 2,921 | 40,655 | 393 | sixth for "wine online" |
Majestic is the clearest case: 61 of its 100 best keywords land on a category. "Champagne" (49,500 searches), "red wine" (18,100), "crémant" (14,800), "barolo" (8,100), "chablis" (6,600): all in first position, all with a dedicated category page, none with an article. The Body Shop follows the same pattern on its non-branded searches: "body butter" (6,600) first, "shower gel" (9,900) second, "face masks" third, all on /collections/ pages.
Hence the rule for site architecture: every group of searches with its own intent deserves a category or subcategory with its own URL, title and copy. Majestic has a page for "crémant" (14,800 searches, first position) and one for "picpoul de pinet" (6,600, first), not a filter inside "sparkling wine" or "white wine". Keyword research for an ecommerce site is 90% this job: deciding which attributes become pages and which stay as filters.
Why indexing everything does not help you rank
The clearest example we have measured is Italian. The electronics retailer ePrice has 80,672 organic keywords on Google Italy, seven times more than the leading Italian wine shop. Of these, 152 are in first position and 11,987 in the top 10. The rest sit lower: 25,504 between positions 11 and 20, 21,845 between 21 and 30. Almost 47,400 keywords on pages two and three, where traffic is close to zero (DataForSEO, Google Italy, August 2026).
The reason shows in the URLs: 73 of ePrice's 100 highest-traffic pages are /pr/ pages, internal search results made indexable. They cover everything and win on little: a page generated by an internal query rarely has the copy, hierarchy and internal links that would make Google prefer it to a competitor's curated category.
This is the faceted navigation trap. A shop with 2,000 products, 12 filters and a sort option produces tens of thousands of possible URLs; Google crawls some of them, indexes fewer and visits the pages that sell less often. The rule we use: a filter becomes an indexable page only if the combination has searches of its own (colour + category, material + category); everything else is handled with a canonical to the parent category, parameters blocked in robots.txt, or noindex. On Bosatta, which sells Vespa and Piaggio spare parts in four countries, the work on variants and model filters mattered more than any copy.
A limit of the data: DataForSEO's keyword counts include variants and synonyms, and estimated visits depend on average CTRs. The order of magnitude holds; the decimals do not.
Recommended reading: how to run an SEO audit.
What a product page needs in order to rank
A few things, and almost no ecommerce site has all of them. A description written by the shop, because the supplier's copy is already on fifty other sites. The full product name in the title, with brand and variant, because that is what people search for: Majestic is first for "moet et chandon imperial brut" (4,400 searches a month) with the product page, and it is the exception that shows how much categories weigh, since most individual product pages catch far smaller queries. Stock and price kept up to date in the code as well, with Product schema markup, which Google uses for rich results and for the product blocks.
Then reviews, which here count as content. In the "wine online" SERP the "People also ask" questions include "What is the best online wine shop?" and "Where is the best place to buy wine online in the UK?": trust is an explicit search intent. Reviews of the product and of the shop (Trustpilot, Google), an "about us" page with address, company number and VAT number, a delivery and returns page: Google and AI models read these to decide whether a shop really exists. We cover this in what E-E-A-T is.
Finally, variants. A product in six sizes and four colours does not need twenty-four indexed URLs: one is enough, with selectable variants and a consistent canonical. The exceptions are variants people search for by name, and those should be decided with search data.
Recommended reading: optimising images for SEO, which on an ecommerce site make up half the page weight.
How much the brand weighs, and what to do if you do not have one
"Body shop" gets 165,000 searches a month. In the sample of The Body Shop's 100 best keywords, 47 contain the brand and account for 78% of the traffic (206,994 visits out of 265,525). Majestic also has 165,000 searches for "majestic wine", but only 6 of its 100 best keywords contain the brand, worth 37% of the traffic.
Two models. The Body Shop lives on awareness built in its high-street shops: 8 of its 100 keywords are store searches ("body shop near me", "body shop glasgow") served by the store locator. Majestic has shops too, but lives much more on its catalogue.
The common belief in the industry is that an ecommerce site without a brand cannot compete in organic search. The data says otherwise: the "wine online" SERP includes reservewines.co.uk and oakhamwinesonline.co.uk, whose brand searches (1,300 and 170 a month) are a tiny fraction of Majestic's, and the skincare SERP includes small producers alongside the big names. They win on precise categories with curated pages. A brand shortens the timeline and lifts the conversion rate; it is not a condition for ranking.
Does an ecommerce blog help you sell?
15 of Majestic's 100 highest-traffic keywords land on guides and blog posts. The guide to how many glasses there are in a bottle of wine is first for a query with 5,400 searches a month; the mulled wine guide is third for a query family with 22,200. Those are volumes comparable to its large categories.
The point is what you do with those visits. Someone checking how many glasses are in a bottle is not necessarily ordering a case tonight. That traffic is for building remarketing audiences, giving Google a reason to see you as an authority on the topic and earning links. Expecting it to convert like a category is the mistake that makes the blog look "useless" to anyone who only looks at revenue per page.
What to write, in this order: buying guides within the category ("how to choose a crémant", "body butter or body lotion: the difference"), which answer the "People also ask" questions and lead to products; pages for brands and producers; recipes and general guides only after that. The copy at the bottom of the category page, the part nobody reads, still carries weight: 150 to 300 words written by someone who knows the products are enough.
AI Overviews have been live in the UK since August 2024, and they now appear at the top of both commercial SERPs we read, naming retailers and linking to them (Majestic, Laithwaites, Naked Wines and Wine Online for "wine online"). Generic informational questions will lose clicks; for category searches the question becomes whether you are among the shops the overview cites. Guides should be written to be cited as a source: see how AI Overviews work.
Recommended reading: creating content with SEO in mind.
Shopify, WooCommerce or PrestaShop: what changes for SEO
The question always comes up: in the UK "shopify seo" gets 480 searches a month, "prestashop seo" 20, "shopify seo agency" 720. The short answer: you can do SEO well, or badly, on all three. The Body Shop's UK site uses Shopify's URL pattern (/collections/, /products/, /pages/, /blogs/ are the platform's signature) and has 2,418 keywords in the top 10. The constraints change, not the possible result.
Shopify. Fixed URL structure (/collections/, /products/, /pages/, /blogs/): no folders like /wine/red/tuscany/. For every product it generates a duplicate URL of the form /collections/category/products/product, with a canonical already pointing to /products/. The robots.txt file has been editable since 2021. Tag and vendor pages need checking: depending on the theme they may be indexable. Speed is good out of the box and gets worse with apps. For foreign markets Shopify Markets handles subfolders or domains per country with automatic hreflang: this is what we use on Bosatta for IT, ES, FR and DE. Scovaventi and Bosatta both run on Shopify.
WooCommerce. Full control over URLs, templates and schema with WordPress plugins (Yoast or Rank Math). The flip side: performance and security depend on hosting and plugins, and faceted navigation generates endless URLs if you do not build it carefully.
PrestaShop. Built multilingual and multi-currency from the start. The layered navigation module produces parameter URLs for every filter combination: without canonicals and robots.txt rules you get the ePrice problem on a small scale.
Below 5,000 products and with a small team, Shopify today has fewer ways to break. If you also sell on Amazon, SEO there is a different discipline ("amazon seo", 260 searches a month in the UK): internal ranking depends on sales, reviews and the words in the title and bullet points, and none of the rules in this article carry over.
How to earn links and authority with a catalogue
Majestic has 21,593 organic keywords and an estimated 637,851 visits a month; Naked Wines, selling wine online to the same country, has 2,921 keywords and 40,655 visits. Part of that gap is catalogue, shops and business model. Much of the rest, in our experience, comes down to years online and links.
For an ecommerce site there are three realistic sources of links. Editorial lists (the Marie Claire article in the skincare SERP and the "best UK natural skin care brands" post next to it both link to the brands they name): you get into them with an offer worth citing and by writing to the editors. The producers and brands you sell: many have a "where to buy" page and update it on request. Content built on your own data (best-selling wines by region, average prices by vintage), which blogs and publications pick up more than any press release.
Links bought in bundles and generic directories do not work. The method is in how to do link building; for internal links, which on an ecommerce site pass authority from strong categories to new ones, there is a dedicated guide.
How to measure whether ecommerce SEO is paying off
With three reports, read together every month.
- Search Console, filtered by folder. Clicks and impressions for /collections/ separated from /products/ and from the blog. If only the blog is growing, you are bringing in readers, not buyers.
- GA4, revenue by organic landing page. A category with 3,000 visits and 2 orders has an intent or pricing problem, not an SEO problem.
- The Search Console "Pages" report. How many pages are indexed and how many are excluded for canonical, noindex or duplication. A shop with 2,000 products and 40,000 indexed URLs is the ePrice case in miniature.
The final KPI is one: organic revenue compared with the same month a year earlier, because seasonality distorts any month-on-month comparison. "Wine delivery" goes from 2,400 searches a month over the summer to 6,600 in December; "online wine shop" from 320 in August to 720 in December (DataForSEO, Google UK, September 2026). A 40% rise from October to November is nobody's achievement. On Scovaventi and Bosatta the number we present to the client is revenue by channel, not rankings. The other indicators are in which SEO KPIs to track.
SEO or Google Ads for an ecommerce site?
Both, but not at the same time and not on the same keywords. Paid clicks on commercial wine keywords cost between £2.08 ("online wine shop") and £5.29 ("champagne online"), with top-of-page bids reaching £4.08 to £4.23 on "buy wine online" and "wine online" (DataForSEO, Google UK, September 2026; CPC converted from USD at £1 = $1.3242). On natural skincare the CPC is lower, £1.63 to £2.24, but the variants with "online" get about 10 searches a month: generic Search has little volume to scale on.
That is why on a new shop we almost always start with Shopping and Performance Max on the catalogue, where CPCs are lower, and in parallel build the categories for organic search, which takes six to twelve months to pay back. On Scovaventi, in one year, €20,000 of Google and Meta budget produced €77,000 in sales with an average basket of €190. On Bosatta, €65,000 across four markets produced €240,000 in sales and 127 dealer enquiries. Every category that reaches the first page is a category on which we can lower the bid the following year.
How to set up the paid side (feed, structure by margin, minimum budget and tracking) is in Google Ads for ecommerce. How to make the two channels work together is in how to make SEO and Google Ads work together.
Frequently asked questions
How much does ecommerce SEO cost?
It depends on how many categories need building and on the technical state of the site: auditing and restructuring a medium-sized catalogue is a project of a few months, followed by monthly work on content and links. Price ranges are in how much SEO costs.
How long does it take to see results?
Technical fixes (canonicals, filters, duplicates) show up in Search Console within 4 to 8 weeks. New categories on keywords with a few hundred searches reach the first page in 3 to 6 months on a domain with a few years of history; those with tens of thousands of searches, such as "champagne" or "shower gel", take years and links. Timelines are in how long SEO takes to deliver results.
Is Shopify bad for SEO?
No. The Body Shop's UK site, built on Shopify's URL structure, ranks first for "body butter" and in the top three for "shower gel" and "face masks". The limits are the fixed URL structure and the tag and vendor pages that need checking. They do not stop you ranking; they mean you need to set up canonicals and robots.txt properly from the start.
What is the best ecommerce platform for SEO?
There is no single best one: Shopify for catalogues up to a few thousand products and small teams, WooCommerce when you need full control and someone maintains WordPress, PrestaShop or Magento for large, multi-country catalogues with a dedicated developer. How you handle categories and filters matters more than the platform.
Is it worth doing SEO on Amazon too?
If you sell on Amazon, yes, but with different rules: internal ranking rewards sales, reviews and the words in the title and bullet points. It does not help your own site rank on Google and, in the niche SERPs we read, Amazon only appears in the product blocks.
Where to start
Open Search Console, go to "Pages" and compare the indexed URLs with the number of products and categories you actually have. If the first is more than double the second, the first job is to close off the filters, not to write copy. If it is lower, you have categories Google cannot find: check the links from the menu and the homepage.
Then take your ten best-selling categories and search Google for the generic name of each one, without "online". If you are not on the first page for any of them, you have your work plan for the next twelve months. If you want to compare your shop's data with the figures in this article, write to us.
