Audit the path from discovery and crawling to indexing, product understanding and mobile use, then prioritise fixes by customer and search impact.
An ecommerce technical SEO audit checks whether search engines and buyers can reach, render, understand and use the pages that matter. It does not replace product truth, category strategy or useful content. A technically indexable thin page is still a weak page.
Use this checklist with the AI-ready website guide, the product-page SEO guide and the ecommerce launch checklist. Assign owners before collecting hundreds of issues.
Set audit scope and business questions
Define the domain, subdomains, countries/languages, platform, catalogue size, recent migrations, peak periods and business-critical templates. Identify priority categories, products and conversion paths.
| Audit scope | Question | Evidence |
|---|---|---|
| Domain and protocol | Which host and HTTPS version are canonical? | DNS, redirects and Search Console |
| Page types | Which templates should be indexable? | URL inventory and CMS |
| Catalogue | How many active products and variants exist? | Product master |
| Markets/languages | Which versions serve which users? | Business plan and hreflang implementation |
| Recent changes | What moved, launched or broke? | Release and migration log |
| Search state | What is crawled, indexed and receiving impressions? | Search Console and analytics |
| Revenue priority | Which pages drive qualified demand? | Reconciled ecommerce data |
Freeze a timestamped baseline. Technical data changes while the audit is running.
Build a trustworthy URL and status baseline
| Source | What it reveals | Limitation |
|---|---|---|
| CMS/product export | Expected pages and business status | May include drafts or stale products |
| XML sitemaps | URLs intentionally submitted | Submission does not prove indexing |
| Search Console | Google-observed indexing and performance | Sampled/aggregated views |
| Crawler | Internal links, status, canonicals and tags | May not render every JavaScript path |
| Server/CDN logs | Actual crawler and user requests | Needs clean bot and path analysis |
| Analytics | Visited pages and journeys | Cannot show unvisited orphan pages |
| Merchant Center/feed | Submitted products and diagnostics | Feed is not site navigation |
Reconcile sources by canonical URL and page type. A URL missing from one list can be the most important finding.
Audit crawlability and access
Google’s technical guidance distinguishes crawling from indexing. Robots rules control crawling; a blocked URL can still be known through links. Use the right control for the intended outcome.
| Check | Pass condition | Failure |
|---|---|---|
| HTTP status | Important pages return stable 200; redirects are intentional | Soft 404, loop, chain or 5xx |
| robots.txt | Blocks only intended paths/resources | Blocks products, CSS or JavaScript needed for rendering |
| Meta/X-Robots | Indexable pages are not noindex | Template leak or conflicting header |
| Internal links | Important URLs use crawlable links | Button/search-only discovery |
| Sitemap | Canonical index-worthy URLs only | Redirects, 404s, parameters or noindex URLs |
| Authentication | Public commerce pages accessible anonymously | Bot sees login or challenge |
| Resources | Critical assets are fetchable | Rendered content disappears |
Do not use robots.txt as a canonicalization method. Google’s current technical SEO guidance and canonical documentation explain these separate controls.
Audit indexation by page type and intent
Compare expected indexable URLs with Search Console’s Page Indexing reports and sample URL Inspection. Group causes instead of fixing pages one by one.
| Page group | Expected state | Investigation |
|---|---|---|
| Homepage and key categories | Canonical and indexable | Access, content, internal links and canonical |
| Active product pages | Indexable when useful and available under policy | Variant/canonical, content and stock state |
| Internal search results | Usually not index targets | Robots/noindex and internal linking |
| Cart, account and checkout | Not search landing pages | Noindex/access and crawl traps |
| Filter/sort states | Controlled deliberately | Parameter explosion and duplicates |
| Expired campaigns | Redirect, archive or remove by user need | Soft 404 and stale sitemap |
| Discontinued products | Case-specific useful alternative or retirement | Blank page or unrelated redirect |
A discovered-not-indexed page may have technical, quality, duplication or prioritisation causes. Do not promise that resubmission alone will fix it.
Control canonicals, parameters and faceted navigation
Ecommerce creates duplicate or near-duplicate URLs through filters, sort, tracking, session parameters and variants. Google describes redirects, rel="canonical" and sitemap inclusion as different-strength canonical signals. Keep signals aligned.
| URL pattern | Default question | Control options |
|---|---|---|
| Tracking parameters | Does content stay identical? | Canonical and clean internal links |
| Sort order | Same products reordered? | Avoid indexation and crawl waste |
| Filters | Is the combination a durable useful landing page? | Controlled index set; manage crawl space |
| Variants | Does each variant need separate search ownership? | Consistent product/variant canonical strategy |
| Pagination | Can later items be crawled and users navigate? | Sequential links and correct page canonicals |
| Print/share views | Is there a duplicate page? | Canonical or remove unnecessary endpoint |
| Uppercase/trailing slash/protocol | Which form is canonical? | One redirect and internal-link standard |
Do not canonical every paginated page to page one. Google’s current pagination guidance says pages in a paginated sequence need their own URLs and sequential crawlable links.
Check ecommerce site structure and orphan pages
Google’s ecommerce structure guidance recommends links from menus to categories, categories to subcategories and then products. Search boxes are not a substitute because crawlers generally do not submit searches.
| Structure check | Pass | Repair |
|---|---|---|
| Menu and hubs | Priority categories reachable logically | Simplify and add descriptive links |
| Categories | Useful grouping with crawlable products | Repair empty/thin or script-only grids |
| Products | Reachable from at least one relevant route | Add category, guide or related-product link |
| Breadcrumbs | Reflect useful hierarchy | Fix inconsistent or non-canonical targets |
| Pagination/load more | Later items have crawlable URLs | Add href-based page sequence |
| Orphans | No important index page lacks internal links | Connect or retire |
| Depth | Priority pages are not buried without reason | Promote through relevant hubs |
Use the content marketing system for helpful guide-to-category/product links. Internal linking should reflect buyer relationships, not arbitrary page-count goals.
Audit product pages, feeds and structured data
| Signal | Consistency requirement | Test |
|---|---|---|
| Title/H1 | Exact product identity and query intent | No template duplication or wrong variant |
| Visible data | Price, availability, variant and specifications | Compare with product master |
| Product JSON-LD | Matches visible purchasable item | Rich Results Test and rendered source |
| Merchant Center feed | Identifiers, price, stock and URL align | Diagnostics and landing-page check |
| Images | Exact variant, crawlable and responsive | Use channel image workflow |
| Canonical | Points to intended product/variant URL | Inspect source and rendered HTML |
| Reviews | Authentic and appropriately marked | No self-serving or hidden mismatch |
Google’s Product structured-data documentation distinguishes product snippets from merchant listings and recommends accurate product information. Structured data does not fix an inaccurate product page. Use the product-page copy template for complete visible decision content.
Check rendering, mobile experience and Core Web Vitals
Test real templates with field data and representative devices. Current Core Web Vitals are LCP, INP and CLS. A passing lab test is not proof that all users or templates pass.
| Area | Audit | Common cause |
|---|---|---|
| LCP | Largest meaningful content load by template | Oversized hero/product image or slow server |
| INP | Interaction responsiveness | Heavy scripts, filters or third-party tags |
| CLS | Unexpected layout movement | Images, banners or fonts without reserved space |
| JavaScript rendering | Critical content and links appear in rendered HTML | Client error or delayed API |
| Mobile controls | Menu, filters, variants and checkout work | Overlays, tiny targets or lost state |
| Accessibility | Keyboard, labels, headings and contrast | Custom widgets without semantics |
| Caching/CDN | Correct content served and refreshed | Stale price, schema or canonical |
Performance fixes must preserve product truth and usability. Compressing away critical detail or deferring a required variant selector is not a successful optimization.
Prioritise fixes by impact, confidence and effort
| Severity | Example | Response |
|---|---|---|
| Critical | Important site section blocked, deindexed or unavailable | Contain and repair immediately |
| High | Canonical/facet issue affects many valuable pages | Template-level fix and monitored rollout |
| Medium | Limited structured-data or internal-link gap | Schedule with clear validation |
| Low | Minor metadata or isolated redirect improvement | Bundle with maintenance |
| Observation | Tool warning without demonstrated problem | Investigate before changing production |
For each issue, record affected pattern, sample URLs, evidence, expected outcome, owner, fix, test, rollback and monitoring date. Test on staging where possible, deploy a limited sample, then re-crawl and inspect live output.
- Fix access, status and canonical conflicts first.
- Repair high-value internal discovery and sitemap quality.
- Control faceted crawl spaces and duplicate signals.
- Align product pages, feeds and structured data.
- Improve rendering, mobile tasks and performance.
- Re-run the same baseline measures after deployment.
- Separate fixed, validated, monitoring and rejected findings.
- Schedule recurring template and release checks.
The audit is complete only when high-priority fixes are validated on the live site and ownership exists for the remaining queue.
Frequently asked questions
What is an ecommerce technical SEO audit?
It is a structured check of access, crawling, indexing, canonicals, URL patterns, site structure, product signals, rendering and performance.
How often should an ecommerce technical SEO audit be done?
Run a full audit after launches or migrations and on a regular cadence, with automated or sampled checks after every material release.
What should be checked first in ecommerce SEO?
Start with HTTP/access, robots and noindex controls, canonicals, sitemap quality, internal discovery and the index status of high-value categories and products.
How should filter URLs be handled for SEO?
Define which combinations have durable search and buyer value, control the rest deliberately, and align crawling, indexation, canonicals, sitemaps and internal links.
Should paginated pages canonicalize to page one?
No. Google advises that paginated pages have their own URLs and sequential crawlable links rather than all canonicalizing to the first page.
How do I prioritise technical SEO fixes?
Use affected business value, severity, scale, evidence confidence, repair effort and rollback risk, then validate the live result instead of closing tickets on deployment alone.
Leave a Reply