The Shopify SEO Checklist for Big Catalogs (2026): 60 Checks for Stores With 500+ SKUs

Most Shopify SEO checklists are written for a store with 40 products. Past 500 SKUs, the problems change. You stop worrying about whether a page has an H1 and start worrying about crawl waste, duplicate URL paths, thin tag pages, orphaned products and a sitemap that says every product changed today.

This is the list we work through on big Shopify catalogs. There are 60 checks, grouped the way we audit: crawl first, then indexing, then the pages that make money. Every platform behavior below was checked against Shopify's documentation and live stores (including Shopify's own Dawn demo store) in September 2026. Where your theme or apps can change the result, we tell you how to check instead of assuming.

What you need before you start#

• Google Search Console with the Page indexing, Sitemaps, Core Web Vitals and Crawl stats reports.
• A crawler (Screaming Frog, Sitebulb or similar) that can crawl the whole store, not a 500-URL sample.
• Theme code access (Online Store > Themes > Edit code) and a staff account that can see URL redirects.
• A terminal for quick curl checks. View-source beats any browser extension for canonicals.

Crawl and robots.txt (checks 1–7)#

• 1. Read your live /robots.txt. Shopify's generated file changed in 2026. Some stores we checked now serve a new commented version that no longer disallows /search or /policies/; others still serve the older rules. Both keep the rules for sort_by, tag combinations (+) and multi-filter URLs.
• 2. If robots.txt.liquid exists, confirm it loops robots.default_groups. A template with pasted plain-text rules freezes an old copy of Shopify's defaults. Shopify's own docs recommend keeping the Liquid objects because "the default rules are updated regularly".
• 3. Confirm nothing valuable is blocked. Test one product, one collection, one blog post and one page with URL Inspection in Search Console.
• 4. Decide what happens to internal search. If /search is crawlable on your store, add a noindex to the search template. The snippet is in our duplicate content guide.
• 5. Storefront password is off. A password-protected store blocks crawlers and the sitemap. It sounds obvious; we still find it on "soft-launched" regional stores.
• 6. Product links exist in the HTML. Crawl with JavaScript rendering off. Collection grids should contain plain <a href="/products/..."> links, not links injected later by an app.
• 7. Group your crawl by URL pattern. Count /products/, /collections/*/products/, ?variant=, /collections/x/tag, ?page=, ?filter., ?sort_by= and /collections/vendors. Large numbers outside products and collections are crawl waste.

Indexing and canonicals (checks 8–15)#

• 8. Compare indexed to submitted. In the Page indexing report, filter by your sitemap. Our rule of thumb: if fewer than about 80% of submitted products are indexed, find out which ones and why before writing any content.
• 9. Read "Duplicate, Google chose different canonical than user". Export the examples and look for a pattern (variant URLs, market subfolders, near-identical products).
• 10. Collection-path product URLs canonicalize to /products/. On a standard theme, /collections/bags/products/tote carries a canonical to /products/tote. Headless builds and heavily customized themes sometimes break this. View source to confirm.
• 11. Collection grids link to /products/handle. Older themes used {{ product.url | within: collection }}, which creates thousands of collection-path links. Search your theme code for within.
• 12. ?variant= URLs canonicalize to the product. They do on standard themes. If an app rewrites your canonical tag, check it again.
• 13. Paginated pages keep their own canonical. Google says don't point page 2 at page 1. On Dawn, ?page=2 is self-canonical, which is correct.
• 14. Tag pages have a decision. /collections/sofas/leather is indexable and self-canonical by default. Index it only if it has unique copy and demand; otherwise noindex, follow.
• 15. Vendor and type pages have a decision. /collections/vendors?q=Brand is also indexable by default. For multi-brand retailers, a real brand collection is almost always better.

Sitemaps (checks 16–21)#

• 16. Submit /sitemap.xml once per domain. Shopify generates one per domain if you run international domains, and each needs submitting.
• 17. Know what's inside. Products (with the primary image), collections, pages, blogs, metaobject pages and, since 2026, sitemap_agentic_discovery.xml. Product files in the big stores we checked held 1,000 URLs each.
• 18. Keep junk out. Set the product status to Unlisted, or use the seo.hidden metafield (integer, value 1), for bundle components, warranty add-ons and test products.
• 19. Don't read meaning into product lastmod. In every product sitemap we pulled on the same day, every product carried that day's date. Google only uses lastmod when it's "consistently and verifiably" accurate.
• 20. Check the agentic sitemap and /agents.md. Open both and make sure the store name and endpoints are right. Full breakdown in the Shopify sitemap guide.
• 21. Submit to Bing too. Bing Webmaster Tools accepts the same sitemap. On one of our client stores, Bing drove $25,000+ of $150,000+ organic revenue in four months.

Collections (checks 22–31)#

• 22. One collection per buyer keyword. A sofa is also a couch, sleeper sofa, sectional and chesterfield. Each is a separate search with separate intent and needs its own collection page.
• 23. Unique title tag and meta description on every indexable collection. Blank SEO fields fall back to the collection title and description, which is rarely what you'd write.
• 24. Copy above and below the grid. A short intro above the products, a longer buying-guide section below. On high-ticket collections, the section below does the selling.
• 25. No thin collections indexed. A collection with one or two products competes with the product itself. Merge it, fill it, or noindex it.
• 26. Short, stable handles. /collections/outdoor-pizza-ovens, not /collections/copy-of-summer-sale-2. Change a handle only with a redirect (check 37).
• 27. Smart collection conditions still match. New products tagged inconsistently drift out of collections. Review the product counts on your top 50 collections monthly.
• 28. Decide on /collections/all and /collections/frontpage. Both are live URLs. /collections/all lists every product and appears in the collections sitemap, so give it a real title and description or keep it out of the index. frontpage rarely deserves to be indexed.
• 29. Pagination links are crawlable. Page 2, 3 and onward should be plain links. "Load more" buttons need a real ?page= link behind them.
• 30. Filters don't create indexable duplicates. On Dawn, filtered URLs canonicalize to the unfiltered collection and multi-filter URLs are blocked in robots.txt. See faceted navigation for when to build a real collection instead.
• 31. No internal links to sort_by URLs. Sort URLs are blocked in robots.txt, so menus and banners that link to them waste links.

Products and variants (checks 32–40)#

• 32. Product titles a buyer would search. No supplier codes, no internal SKUs in the title.
• 33. Descriptions aren't copied. Chinese brands often reuse the Amazon listing on Shopify. Google sees the same text on Amazon, marketplaces and resellers; rewrite the Shopify version.
• 34. Variants vs separate products is a keyword decision. If "black leather sofa" has its own demand, a separate product (or combined listing on Plus) can rank where a variant can't. Shopify allows up to 2,048 variants and three options per product.
• 35. Out-of-stock products stay live. Keep the page, show alternatives. Only redirect a product when it's discontinued for good.
• 36. Alt text on product images. Describe the product and the angle, not a keyword list.
• 37. Handle changes create redirects. When you edit a handle in the admin, keep "Create a URL redirect" ticked. Bulk handle changes through apps need the same option.
• 38. No duplicate products. The same item listed twice to sit in two collections splits reviews, links and rankings. Use one product in two collections.
• 39. Reviews render in the HTML. If the review text only loads in an iframe or after interaction, crawlers may not see it.
• 40. Product category is set. Shopify's standard product taxonomy feeds filters, smart collection conditions and Shopify Catalog for AI channels.

Structured data and feeds (checks 41–46)#

• 41. One Product block per page. Dawn outputs {{ product | structured_data }}, which becomes Product or ProductGroup (for products with variants). Review apps often add a second. Keep one source of truth.
• 42. Offers match the page. Price, currency and availability in the JSON-LD must match what shoppers see, per market.
• 43. Organization and WebSite markup on the home page. Name, logo, sameAs links to real profiles.
• 44. BreadcrumbList on products and collections.
• 45. Merchant Center has no price or availability mismatches. Mismatches get products disapproved in Google's free listings and Shopping ads. Fix feed errors weekly.
• 46. Run one URL per template through the Rich Results Test. Product, collection, article, page. See our schema service for what we add beyond the defaults.

Internal linking (checks 47–52)#

• 47. Top collections are in the main navigation HTML.
• 48. Collections link to related collections. "Built-in grills" should link to "grill islands" and "outdoor kitchen cabinets".
• 49. Buying guides link to collections, not only products. A guide that links to a collection passes value to 40 products at once.
• 50. Products link back to their collection. Breadcrumbs, "more from this collection" blocks.
• 51. No internal links to redirected URLs. Menus, rich-text descriptions and old blog posts are the usual culprits.
• 52. No orphan products. Every product in the sitemap should be linked from at least one crawlable collection page.

International (checks 53–56)#

• 53. hreflang is present and correct. Shopify adds hreflang tags for Markets domains and subfolders. View source on a product in each market and confirm the set.
• 54. Translated pages are really translated. Handles, titles and meta descriptions in /de/ should be German, not English.
• 55. No forced IP redirects. Googlebot mostly crawls from the US. Suggest a market with a banner or selector; don't force it.
• 56. Redirects cover subfolders. Shopify applies a redirect to every market subfolder automatically, but a market-specific destination needs its own redirect. More in our international SEO service.

Speed and monitoring (checks 57–60)#

• 57. Core Web Vitals pass on collection templates. Good means LCP ≤ 2.5 s, INP ≤ 200 ms and CLS ≤ 0.1 at the 75th percentile. Check Search Console's Core Web Vitals report grouped by template.
• 58. Removed apps left no code behind. Search theme.liquid and snippets for scripts from apps you uninstalled.
• 59. Only above-the-fold images load eagerly. The first row of a collection grid should load immediately; the rest can lazy-load.
• 60. A monthly monitoring routine exists. Page indexing report, new 404s, the Crawl stats report and the top 50 collections' positions. Put it on a calendar.

How to prioritize 60 checks#

The checks on crawl, canonicals and sitemaps are the "do no harm" layer. The revenue comes from checks 22 to 24 and 47 to 50: more collection pages targeting real buyer keywords, better copy on them, and internal links that push authority into them. That's the work behind our technical SEO and large catalog SEO services.

Look on page one before you build anything. Google shows you exactly what it rewards.