I run SEO for online stores: category and product architecture, faceted-navigation control, product schema, and content that ranks for buying queries — for D2C brands and multi-category retailers in India, the UAE and the UK.
Why e-commerce SEO fails differently
Ordinary SEO problems are about content quality. E-commerce problems are about scale and structure, and they have their own specific pathologies:
Faceted navigation generating infinite URLs. Colour, size, price and brand filters combine into tens of thousands of near-identical pages, and the crawler spends its budget on permutations rather than on the pages that sell.
Category pages with forty words. They are the highest-intent pages on the site — someone searching “22K gold jhumka” wants a category, not a blog post — and they typically carry less text than a tweet while competing against marketplaces with thousands.
Out-of-stock pages 404ing. Years of accumulated ranking signal deleted because a product went out of season. Discontinued is not the same as gone; the successor product or the parent category is a redirect target.
A blog about “5 summer trends” that ranks for nothing anyone buys, commissioned because someone said content matters, disconnected from any commercial query.
Meanwhile the marketplaces eat the generic terms. Which is fine — the win is elsewhere.
Where the winnable demand actually is
Specific buying queries. “Jewellery” browses; “22K gold jhumka latest design price” buys. The long, precise query where the buyer has already decided what they want and is choosing where to get it — that is the query a well-built category or product page wins, and it converts several times better than the head term you would have lost anyway.
That reframing changes what the work is. Instead of fighting Amazon for a word, you build the architecture that captures a thousand specific intents, each small, collectively larger than the head term, and each far closer to a purchase.
What the engagement covers
- Architecture. Categories that map to how people actually shop rather than to how your warehouse is organised, with facets crawl-controlled so search engines index intentions rather than permutations.
- Category page depth. Enough real content to compete, structured so it helps a buyer choose rather than padding for a word count.
- Product templates. Title patterns, specifications as structured data (Product, Offer, Review), honest availability handling — discontinued products redirect to successors rather than dying.
- Buying-intent content. Guides that end in your products, built around queries people search before purchase, rather than trend pieces that rank for nothing.
- Marketplace coexistence. An explicit decision about what to concede and what to contest, so budget goes where it can win.
- Migration protection. The redirect map before any replatform, because that is the number-one killer of e-commerce organic traffic and it is entirely avoidable.
The number I ask for first
Your duplicate URL count. Run a crawl, count how many indexable URLs exist versus how many products and categories you actually sell. In most stores I audit the ratio is somewhere between five and fifty to one, and every one of those extra URLs is consuming crawl budget that should be spent on pages that convert.
It is never a small number, and it is the fastest thing to fix. Nothing else in this service produces as much movement for as little work as bringing the crawlable surface back in line with the sellable one.
How the engagement is sequenced
Weeks one and two: the crawl and the architecture verdict. Full crawl, duplicate-URL count, facet analysis, and a decision on category structure. This phase produces the map: which pages carry commercial weight, which are noise, and which intentions have no page at all.
Weeks three to six: structural fixes. Facet controls, canonical discipline, redirect cleanup for discontinued products, product schema, and the templates that make every future product page correct by default. Most of the technical recovery lands here, and it is measurable in crawl-budget terms within weeks.
Weeks six onward: category depth and buying-intent content. The slower half, sequenced by commercial value rather than by search volume — your highest-margin categories first, whatever the keyword tool says about traffic.
Throughout: measurement in Search Console on queries and pages rather than in a vendor dashboard, because that is the only ground truth about what buyers actually typed.
What I will tell you not to do
Do not build a page for every keyword variation. Do not write four hundred words of filler under every category because a competitor did. Do not chase the head term your marketplace competitors own. And do not replatform without a redirect map, ever — that single mistake has cost stores more organic revenue than every other item on this page combined.
Why category pages deserve most of the budget
A product page can only ever rank for one product. A category page ranks for the pattern — the material, the occasion, the price band, the use case — and it keeps ranking as individual products come and go beneath it.
That makes categories the compounding asset in an e-commerce site, and they are almost universally under-invested because they feel like navigation rather than content. The stores that win organic in competitive categories are the ones that treated their category pages as the primary selling surface rather than as a grid with a heading.
The practical consequence for budget: one well-built category page usually outperforms twenty optimised product pages, and it keeps doing so after the products beneath it have been replaced twice over.
The three numbers I report monthly
Indexable URLs versus sellable products. The ratio that reveals whether crawl budget is being spent on your catalogue or on filter permutations. It should fall steadily through the engagement and then stay flat.
Category-page rankings for buying queries, not head terms. The specific, purchase-adjacent phrases where you can genuinely win, tracked as a set rather than cherry-picked.
Organic revenue by landing-page type — category, product, guide — from your own analytics rather than an estimate. This is the number that decides where the next month’s effort goes, and it frequently contradicts what the keyword tools recommend.
Three numbers, one page, monthly. If a report needs more than that to justify itself, it is decorating rather than informing.
What this costs you in effort
Realistically: a few hours of your merchandising team’s time in the first fortnight, because category structure is a commercial decision rather than a technical one and I should not be making it alone. Then product-template review, then sign-off on the content sequence.
After that it is largely mine to run. But the architecture phase genuinely needs someone who knows why your customers group things the way they do — and that person is never the SEO consultant.
Related reading
/services/technical-seo — the foundation underneath · /services/shopify-development — the platform-side work · /blog/technical-seo — the audit method · /industries/retail — the trade context
Send the store URL
The first call includes the duplicate-URL count and the three biggest structural findings, free. If your team already knows the number, we will skip straight to what to do about it. /contact — paisa vasool marathon, not a sprint.
Questions I actually get
Shopify, WooCommerce, Magento or custom?
All of them. The crawl does not care about your platform; the template mechanics and the default traps differ, and I know each platform's particular way of generating duplicate URLs. Shopify's collection-and-tag structure is the most commonly misconfigured of the three.
We spend heavily on ads — why bother with organic?
Because paid customer-acquisition cost only goes up and organic compounds. That is the boring answer and it remains true. The sharper version: ads stop the day you stop paying, and a category page that ranks keeps selling through the months when the ad budget is cut.
How long before it shows?
Technical fixes bite within weeks. Category authority and content compound over months. Anyone promising festival-season results from a September start is selling you optimism — the honest move is starting early enough that the season benefits.
Do you write product descriptions at scale?
Templates and the top pages, yes. At catalogue scale I build the system your team runs, because a thousand descriptions written by an outsider will be worse than a hundred written by people who know the products plus a structure that makes them consistent.
Can we outrank Amazon and the marketplaces?
On generic head terms, usually not, and I will say so rather than take the budget. On specific buying queries — the long, precise ones where the buyer knows what they want — regularly yes, and those convert far better anyway.
What happens during a replatform?
The redirect map is built before launch, not after. Replatforming is the single most common cause of catastrophic e-commerce traffic loss, and it is entirely preventable with a URL inventory and a mapping exercise that takes days.