Ecommerce SEO.
From catalogue
to customer.

1. Priority product, collection and editorial URLs return stable 200 responses.
2. robots.txt does not block pages or resources required for rendering priority templates.
3. No priority templates carry accidental noindex directives.
4. Canonical tags resolve to deliberate, indexable destination URLs.
5. Canonical targets use the preferred protocol, hostname and trailing-slash convention.
6. XML sitemaps contain only canonical, indexable 200 URLs.
7. Product URLs are discoverable through crawlable HTML links, not only internal search or JavaScript events.
8. Pagination exposes crawlable sequential URLs or equivalent crawl paths.
9. Load-more and infinite-scroll implementations have crawlable URL equivalents.
10. Important products are not orphaned when they leave merchandising modules.
11. Parameter and tracking URLs do not create uncontrolled indexable duplicates.
12. Internal search-result URLs have an explicit crawl/indexation policy.
13. Large catalogues have a documented crawl-budget strategy rather than indiscriminate blocking.
14. Server logs or crawl evidence are used to validate bot access for large/complex stores.
15. Indexation is segmented by template and commercial value, not reviewed only as a domain-wide total.

1. Every major commercial query cluster has a clearly assigned preferred page type.
2. Category hierarchy reflects how customers shop, not only the ERP or internal organisation chart.
3. Broad categories link to durable subcategories before products where the taxonomy warrants it.
4. High-value collection pages are reachable within a sensible number of clicks from strong hubs.
5. No two priority collections exist solely because different teams created them for similar intent.
6. Cannibalisation is checked using query-to-page data, not only keyword-tool overlap.
7. Internal search terms are reviewed for missing categories, attributes and synonyms.
8. Site-search no-result queries are treated as product/discovery research.
9. SERP format is checked before deciding whether a query belongs to a PDP, PLP or guide.
10. Category creation requires meaningful demand, inventory depth and durability.
11. Seasonal pages have a reuse/retention strategy instead of being rebuilt every year.
12. Campaign landing pages do not unnecessarily compete with evergreen category URLs.
13. Brand, product-type and use-case architecture is explicit for multi-brand retailers.
14. Editorial content links into commercial hubs where it genuinely advances the buying task.
15. Architecture decisions are documented so merchandising and development do not undo them later.

1. Every facet has a documented crawl/indexation decision.
2. Filter combinations cannot generate an effectively infinite crawl space.
3. Sort-order URLs are prevented from creating duplicate indexable pages.
4. Low-value multi-filter combinations remain shopping tools rather than SEO landing pages.
5. High-demand durable facet combinations are promoted into intentional landing pages where warranted.
6. Indexable filtered experiences have self-referencing canonicals and useful unique value.
7. Indexable facet pages are linked internally rather than dependent on crawlable filter links alone.
8. Zero-result filter states cannot become indexable thin pages.
9. Facet values use a controlled taxonomy rather than inconsistent free text.
10. Filters reflect category-specific decision attributes, not only generic price/brand/colour.
11. Applied filters are visible and removable for users.
12. Multi-select behaviour matches the way customers actually refine the category.
13. Pagination URLs are unique, crawlable and internally linked.
14. Each paginated URL contains the correct products without requiring a user action to reveal them.
15. Products beyond page one remain discoverable without relying solely on sitemaps.

1. Each priority PLP has a clear H1 aligned with the product set and search intent.
2. Title tags differentiate category scope and avoid boilerplate duplication.
3. Intro copy explains the collection rather than padding the page with generic SEO prose.
4. Critical buying information is visible before excessive editorial copy pushes products down.
5. Product cards expose the attributes customers need to compare options.
6. Pricing and availability are accurate on PLP cards where displayed.
7. Filters use terminology customers understand.
8. Category-specific filters exist for the important decision attributes.
9. Products are merchandised so strong/representative inventory appears early without hiding relevance.
10. Related subcategories and adjacent collections are linked contextually.
11. Breadcrumbs reflect the intended hierarchy and are crawlable.
12. Empty categories have a deliberate UX and indexation response.
13. Low-stock categories are monitored before they become poor search destinations.
14. Helpful buying guidance is available where the decision genuinely needs explanation.
15. PLP performance is measured against both organic acquisition and product-finding/conversion outcomes.

1. Product names use shopper language and preserve the real product identity.
2. PDP copy explains what the product is, who it is for and the important constraints.
3. Important specifications are present as text/data, not trapped only in images or PDFs.
4. Variant options expose correct price, availability, imagery and identifiers.
5. Each important variant has a stable deep-linkable state where the platform supports it.
6. Product images are high quality, crawlable and supported by descriptive alt text.
7. PDPs include useful visual evidence such as additional angles, scale, fit or use context where relevant.
8. Shipping expectations are clear before checkout.
9. Return policy information is easy to find and internally consistent.
10. Reviews provide genuine decision support rather than decorative star counts.
11. FAQs answer product-specific objections rather than repeating generic company information.
12. Related products and accessories are contextually useful and crawlable.
13. Comparison/help content is linked where customers commonly hesitate between options.
14. Out-of-stock messaging provides useful next actions rather than a dead end.
15. PDP success is measured in revenue/conversion as well as impressions and rankings.

1. Merchant Center is connected and treated as organic-discovery infrastructure, not only a paid-media feed.
2. Product feed IDs are stable and never casually recycled for different products.
3. GTINs are supplied when valid manufacturer identifiers exist.
4. GTINs are never invented to fill a field.
5. Brand values reflect the actual manufacturer/brand rather than the retailer where inappropriate.
6. Variant items share a stable group identifier.
7. Feed title, landing-page title and selected variant describe the same purchasable item.
8. Feed price matches the landing-page price for the selected offer.
9. Feed availability matches the landing page and backend stock state.
10. Product schema reflects visible on-page truth.
11. ProductGroup/variant markup matches the actual variant model.
12. Shipping and returns data is supplied in the appropriate feed/schema/account location.
13. Product highlights describe useful benefits rather than keyword stuffing.
14. Product detail fields capture technical, verifiable attributes where useful.
15. Feed diagnostics are part of SEO QA and are not left solely to paid media.

1. New products have a planned discovery path before launch.
2. New PDPs are linked from relevant collections immediately when they become commercially useful.
3. Temporarily out-of-stock products remain live when the product is expected to return and the page still serves demand.
4. Temporary OOS pages clearly communicate status and offer relevant alternatives or restock actions.
5. Permanently discontinued products are distinguished from temporarily unavailable products.
6. Discontinued products with a true successor have a deliberate redirect/replacement strategy.
7. Discontinued products without an equivalent are not blindly redirected to unrelated categories.
8. Product removal rules consider traffic, backlinks, demand, replacement and customer utility.
9. Seasonal products have documented off-season handling.
10. Returning seasonal PDPs reuse durable URLs where appropriate.
11. Variant discontinuation does not break the parent product experience.
12. Feeds remove or update discontinued inventory consistently with the website.
13. Sitemaps stop advertising dead inventory as current.
14. Internal links are updated when a product leaves the catalogue permanently.
15. Lifecycle decisions preserve valuable information when the page still satisfies research demand.

1. Content strategy begins with customer decisions, not monthly blog quotas.
2. Buying guides resolve real product-selection questions and link to relevant commercial pages.
3. Comparison content demonstrates criteria and evidence rather than generic summaries.
4. Editorial recommendations disclose relevant commercial relationships.
5. First-hand tests include original observations, images or measurements where possible.
6. Experts or qualified staff are attributed where judgement is central to the recommendation.
7. Internal search vocabulary informs category, synonym and content decisions.
8. On-site search tolerates common misspellings and customer language where the platform allows.
9. No-result searches feed a recurring merchandising/content backlog.
10. Blog/editorial pages do not cannibalise commercial pages by targeting the same task with the wrong format.
11. High-performing informational pages contain deliberate routes to relevant product discovery.
12. Commercial pages link back to deeper guidance where it reduces uncertainty.
13. Content decay is monitored for product availability, outdated models and superseded recommendations.
14. AI-generated copy is human-reviewed for product truth, differentiation and legal accuracy.
15. Automation is used for consistency and scale; important pages receive human judgement.

1. Priority commercial pages earn links/mentions through genuinely useful assets, products or stories.
2. Digital PR is connected to categories/products that have commercial upside, not treated as a disconnected vanity channel.
3. Brand mentions are monitored alongside backlinks.
4. Review acquisition follows platform policies and never relies on fabrication or incentives that breach rules.
5. Review content is visible and useful to customers.
6. Review markup matches reviews actually shown on the page.
7. Product review editorial demonstrates first-hand or expert evidence where claimed.
8. Media coverage reinforces real product claims rather than invented superlatives.
9. Creator/influencer content is considered part of the wider product evidence ecosystem.
10. Manufacturer and retailer product information are consistent where they refer to the same item.
11. Priority brands/products have credible independent references beyond the merchant site.
12. PR/product launches preserve stable URLs and reusable evidence after the campaign ends.
13. Comparison/review queries are monitored for inaccurate third-party narratives.
14. Leadership/expert commentary is used where it creates differentiated category expertise.
15. Authority work is measured against commercial visibility and assisted outcomes, not link count alone.

1. AI shopping is treated as an emerging discovery layer, not a replacement for Google/Search/Shopping fundamentals.
2. Google Merchant Center product data is complete enough to support richer and AI-driven discovery surfaces.
3. OpenAI/Shopify agentic-channel eligibility is understood where commercially relevant.
4. OpenAI product-feed requirements are mapped against the store's current catalogue model where applicable.
5. Product identifiers remain stable across website, feed and agentic integrations.
6. Prices and availability update frequently enough for the volatility of the catalogue.
7. Shipping/returns/product-policy data is machine-readable and human-readable.
8. AI/search monitoring separates brand mention, product citation, recommendation and referral outcomes.
9. Agent-driven journeys are not assumed to be fully visible in standard web analytics.
10. Log-level and platform reporting are considered where first-party referral data is incomplete.
11. Organic revenue is reported by landing-page/product/category cohort where possible.
12. SEO opportunity models incorporate current demand, conversion rate, AOV/margin and paid dependency.
13. Paid and organic overlap is reviewed to identify expensive demand the brand could earn more efficiently.
14. Experiments define the expected commercial mechanism before implementation.
15. The roadmap is prioritised by impact, confidence and effort rather than by audit-error count.

It can include demand and category strategy, technical access and indexation, product-page quality, Merchant Center and structured data, stock-lifecycle rules, decision content and commercial measurement. The scope follows the constraints in your store rather than a fixed monthly task quota.

The approach applies to Shopify, WooCommerce, Adobe Commerce and custom or headless stores. Implementation depends on the platform, integrations and development capacity. We establish those constraints during the review and work with the people responsible for delivery.

Technical corrections and product-data fixes can affect eligibility sooner than competitive growth appears. Outcomes depend on the starting point, demand, competition, stock and implementation speed. The first 90 days establish the foundations and measurement; they are not a guaranteed revenue deadline.

No. A filter is a shopping control first. A permanent search landing page should pass the demand, inventory, differentiation and durability tests. Agree a deliberate crawl, indexation and canonical policy for the remaining combinations.

No blanket rewrite is required. Prioritise accurate facts and useful evidence: fit, compatibility, measurements, original images and product comparisons. Cosmetic uniqueness is a poor substitute for helping a customer decide.

No. Structured data can support understanding and eligibility for supported search features. It does not guarantee rankings, rich results or inclusion in an AI answer.

Yes. The Index Review is the starting point for understanding the constraints, opportunities and decisions before agreeing an ongoing programme. We can then define the work and delivery responsibilities together.

Yes. Ecommerce search often depends on shared ownership. A feed mismatch or template defect needs a clear owner, acceptance criteria and a validation plan across the relevant teams.