As an SEO & GEO expert I help online shops recover revenue where it is usually lost: category architecture, faceted navigation, the internal link graph and a conversion funnel whose economics were never thought through. Traffic is rarely the bottleneck.

“Most shops are fighting their own architecture. Fix the foundation and the rest follows.”
Broken facet URLs, cannibalizing categories, product data without markup, crawl budget lost in filters, Core Web Vitals in the orange, a checkout full of conversion killers. No campaign repairs a broken foundation.
Deliberate facet handling (indexable or not, consistent canonicals, correct internal links) measurably improves session quality and revenue per session, often by double digits. The lever is structure, not creative.
Four layers in every e-commerce project. Without them, no single lever scales.
Signals from four layers: crawl-budget health, category revenue velocity, checkout drop-off and margin attribution. Every KPI traceable to SKU level.
Crawl budget, canonical strategy, faceted-navigation logic, pagination, product markup (Product, Offer, AggregateRating, Review), Core Web Vitals on real devices.
Category copy with intent depth (no SEO fluff), PDPs with structured specifications, merchandising-level FAQs, cross-sells by semantic proximity.
Event-level conversion audit (checkout drop-off, basket abandonment, product friction), an A/B test pipeline with power analysis, mobile-first testing.
Cross-channel revenue attribution, margin-aware assortment management, paid media integrated with SEO. No siloed KPIs, no last-click blind spots.
E-commerce SEO in 2026 is the work of turning a shop's catalogue into search visibility that pays off in revenue, not just rankings. It joins product structured data, faceted-navigation governance, category architecture and conversion optimization into one system. Product, Offer and AggregateRating markup and a clean Merchant Center feed shape how products appear in Google Shopping, organic product results and AI Overviews. Faceted navigation decides where crawl budget goes, so only filter combinations with real search demand should be indexable. Cannibalization between category, filter and product pages is managed with a keyword-to-URL map. Marketplaces such as Amazon, eBay or Google Shopping follow their own ranking logic and need their own playbooks. And conversion optimization is the second half of the revenue equation (traffic × conversion rate × average order value): SEO gains without CRO discipline leak out of the funnel.
Add SKU-level seasonality forecasting, hreflang that links only products that really exist in each market (with correct price, currency and delivery data), and shopping feeds held to Merchant Center standards instead of being mirrored from the PIM. E-commerce SEO sits where technology, assortment, pricing and conversion architecture meet.
Product, Offer, AggregateRating and Review markup in sync with the front end. Variants via ProductGroup or separate PDPs, depending on demand and SKU depth.
Indexing rules per facet combination on hard criteria (volume, result-set size, intent). Canonicals, robots and internal links as one system.
Amazon, Google Shopping, comparison sites and regional marketplaces as disciplines of their own, with pricing discipline and incrementality tracking per channel.
Category, PDP, basket and checkout as one conversion path. Powered A/B tests, mobile-first, lower checkout INP.
Four phases, from the initial audit to an ongoing retainer, each with clear hand-overs and deliverables.
Crawl health, canonicals, facet inventory, markup coverage, cannibalization map, assortment performance per category and SKU, competitor and marketplace snapshot.
Product, Offer, rating, review, shipping and return-policy markup. Facet indexing rules, pagination, hreflang for multi-market shops.
Category copy, PDP optimization, trust signals, filter and sort UX, checkout INP, an A/B test pipeline, mobile-first testing.
Monthly revenue reviews, seasonality forecasts, marketplace management, update response, schema drift monitoring, SEO briefings for new ranges.
All four act on the same revenue KPI but differ in levers, time to impact and risk. SEO builds the long-term foundation, paid media and Shopping ads capture demand fast, marketplaces add reach at a channel cost. Run them against each other and you lose margin.
Each shop type has its own levers. What works in fashion fails in B2B, and vice versa.
Deep SKU ranges with exploding variants, seasonal collections, out-of-stock handling, brand vs. generic conflicts, editorial content for trends, frequent sales.
Login pricing, customer-specific ranges, data sheets as content, long cycles, procurement integration, RFQ funnel next to checkout, low volume but high order values.
Amazon, eBay or Otto as the main channel: sponsored ads, Buy Box dynamics, A+ content, review acquisition, platform ranking factors, pricing discipline versus your shop.
Brand building alongside performance, PDP storytelling, subscriptions, customer lifetime value as KPI, content hubs, an AI citation strategy for generative search.
hreflang architecture, assortment gaps per country, currency and VAT, local payment methods, regulatory disclaimers, regional marketplaces (Allegro, Bol, Cdiscount).
YMYL compliance, prescription vs. OTC separation, indication restrictions, mail-order pharmacy rules, MedicalEntity markup, pharmacist-level E-E-A-T, health-insurance integration.
Audit, implementation project or ongoing retainer. Terms are set in the scoping call, based on shop size, assortment depth, markets and KPI responsibility.
Crawl health, markup coverage, facet inventory, cannibalization map and a prioritized roadmap, plus workshop and executive summary.
Schema rollout, facet refactor, PDP optimization, category content and CRO pipeline set-up, steering your teams and vendors.
Monthly reviews, seasonality forecasting, marketplace management, update response and schema drift monitoring.
As a rule of thumb, from around 5,000 products with three to four active filters, because URL combinations then multiply and dilute PageRank. Test each combination: at least 50 searches a month, at least 10 products, measurable commercial intent. If fewer than two apply, canonicalize to the parent category and set the combination URL to noindex.
Use a split layout: a 200-400 word introduction above the fold for intent coverage, the product grid directly below as the main UX path, and optionally a buyer's guide or FAQ block after the grid. On mobile, collapse the text behind a "Read more" toggle so it doesn't delay LCP.
Separate PDPs when variants have distinct search demand, which is typical in fashion. One PDP with ProductGroup markup when many SKUs share the same intent (tech, tools). A hybrid of master PDP and deep-linkable variant URLs for complex configurations such as furniture or B2B. Decide by the search-volume gap between variant terms and the parent.
LCP is mainly an acquisition lever. INP becomes revenue-critical in the checkout, where lag when choosing shipping, switching payment method or entering a voucher costs completions. Levers: lighter event handlers, isolated state updates, preloaded payment scripts, analytics via the Beacon API. Measure INP per checkout step, not per page.
Three layers: an event stream in a CDP or data warehouse as the single source of truth, identity resolution via user ID and hashed email, and a Markov-chain or Shapley model on the cleaned event graph. Treat last click as one alternative model, not the truth. Server-side tracking (Conversions API, Measurement Protocol) closes Safari and blocker gaps.
It is a multi-disciplinary system. Product, Offer and AggregateRating markup shape visibility in Google Shopping, Merchant Center and AI Overviews, and faceted-navigation governance decides how crawl budget is spent. Cannibalization, seasonality and marketplaces are disciplines of their own. Review, availability and shipping markup change how listings look, and with it their click-through rate.
For every PDP: Product (gtin, mpn, brand), Offer (price, priceCurrency, availability, priceValidUntil, hasMerchantReturnPolicy, shippingDetails), AggregateRating and Review. For categories: ItemList. Optional: ProductGroup, MerchantReturnPolicy, OfferShippingDetails. Markup must match what the page shows, or you risk item disapprovals and suspensions in Merchant Center.
Keep the hero assortment exclusive to your shop, or list it on marketplaces only with strict RRP discipline, and move long-tail stock to marketplaces for liquidity. Never undercut your own shop on price comparison sites. Run Google Shopping as a complement to organic and track incrementality per channel. Amazon needs its own content, A+ and sponsored logic.
Map intent per cluster: the generic term ("running shoes") to the category, a specific term ("women's trail running shoes") to a landing-ready facet URL, brand plus model to the PDP. Keep that keyword-to-URL map as the single source of truth, crawl for cross-ranking conflicts and deindex or canonicalize one URL when two compete.
CRO is the second half of the revenue equation: traffic × conversion × AOV. SEO brings qualified traffic; CRO decides what it earns. Levers: category (filter UX), PDP (trust, shipping, variants), basket (fewer steps, express checkout), checkout (forms, INP, error recovery). Every change is A/B tested with a power analysis.
Link only products and categories that exist in both markets with hreflang, never placeholders. Each market gets its own URLs and structured data (price, currency, availability, shippingDetails). Use 410 for permanently removed SKUs and 200 with a clear "not available" label for temporary gaps. Wrong priceCurrency values cause item disapprovals and, if systematic, can lead to a Merchant Center suspension.
30 minutes. We look at your category structure, conversion path and attribution, and name the three levers with the highest margin impact.