Ecommerce SEO audit: find and prioritise store issues

For Shopify SEO audits and other online stores, check indexation, product pages and search performance before prioritising fixes you can verify.

An ecommerce search engine optimisation (SEO) audit finds which store pages search engines can reach, understand and index. It turns confirmed failures into repairs you can retest. Start with search access before rewriting product descriptions. A blocked product template can affect a whole range while each individual description looks fine.

Set the audit baseline and sample

Define the business symptom and save its evidence before changing the store. An SEO audit is a structured investigation of search access, technical delivery, on-page information and authority signals. Run one after a visibility loss or catalogue change, and before a redesign puts existing search traffic at risk.

Record the storefront, affected period and symptom, such as fewer organic visits to camping-chair products. Compare equivalent periods and separate branded searches from product searches. In Google Search Console, export clicks, impressions, queries and affected pages. In analytics, save organic landing-page sessions, purchases and revenue using consistent attribution settings. Check tracking changes before treating a reporting drop as lost demand.

Ask the owner for read-only Search Console and analytics access. Without it, public responses and page inspection still support a narrower audit. Label the missing account evidence in your report and avoid conclusions about traffic you cannot see.

Build a sample across products, categories or collections, variants, filters, pagination and unavailable items. Include strong pages and weak pages from each shared template. Record the selection method and counts. A comprehensive checklist covers the diagnostic areas. Inspecting a sample does not mean every catalogue URL was checked.

Copyable ecommerce SEO audit checklist

Work down this checklist. Save a result or issue ID for each check and each selected page.

Order and checkRequired evidenceAffected page sampleExpected resultResponsible owner
1. Save the baselineSearch Console and analytics exportsAffected and comparison landing pagesSymptom and period are definedSEO lead and analyst
2. Test search accessResponses, redirects, robots rules and indexing directivesEvery sampled templateIntended public pages are accessible and indexableDeveloper
3. Check canonical and sitemap consistencyCanonical values, sitemap entries and URL InspectionProducts, variants and collectionsSignals identify the intended search pageDeveloper and SEO lead
4. Follow discovery pathsCategory links, pagination and catalogue inventoryDeep products and filtered resultsIntended products have reachable linksMerchandiser and developer
5. Audit on-page informationVisible copy, metadata and product markupOne product and category per templateSearch task and product facts agreeContent and catalogue owners
6. Test rendering and mobile useRaw and rendered output, mobile observations and performance evidenceShared templates and app-dependent contentProduct details load and navigation worksFrontend developer
7. Inspect links and search competitorsInternal links and referring-page samplesBroken destinations and equivalent search tasksDefects have verifiable source evidenceSEO lead
8. Check locales and redesign risksLocale pairs or old-to-new URL mapRegional and changing routesDestinations preserve their intended purposeDeveloper and regional owner
9. Retest released repairsOriginal failed check and new resultAffected pages plus neighbouring templatesFailure is resolved without a new access problemIssue owner and SEO lead

Scale the sample by shared failure points. For an enterprise store, group by template, catalogue depth, locale and deployment dependency, then expand any group with a repeated fault. Keep both the sampled count and total known affected count.

For a business-to-business (B2B) trade catalogue, separate public product information from account-only prices and ordering. Inspect public access without a customer session. For WordPress with WooCommerce, record the active theme, SEO plugin and product-extension versions beside rendered metadata. Compare product, category and attribute templates for conflicting output. A plugin configuration screenshot alone does not show what the public page delivers.

Check access, indexation and search coverage

Check responses first, followed by robots rules, indexability, canonicals and sitemap inclusion. Before judging indexability, confirm that rendered links let a crawler discover the page.

  1. Request the affected URL and follow its redirects. An intended live product must end at its correct public page with a successful response. A redirect loop or server error goes to the developer. A retired item needs the agreed removal or relevant replacement behaviour.
  2. Read the robots rules that apply to the URL and resources it needs. Give an accidental collection block to the developer, with the exact matching rule. The correction must allow the intended public collection to be fetched.
  3. Inspect incoming category links and rendered pagination. If navigation only opens a search interaction, record the missing crawlable product links for the frontend owner.
  4. Inspect indexing directives in both response headers and page markup. An unintended noindex on live products goes to the template owner. Remove it only from pages intended for search, then confirm the released response.
  5. Compare the declared canonical with the intended page. Send conflicting template output to the developer and agree which equivalent URL belongs in search.
  6. Compare sitemap entries with that decision. Give redirected or excluded entries to the sitemap owner. The corrected sitemap must list the intended canonical destinations.

Google’s robots guidance distinguishes crawl control from index exclusion. A robots block can prevent Google from seeing a page’s noindex. Keep private trade content behind access controls.

Use URL Inspection to compare Google’s stored information with a live test. Save the reported crawl outcome and Google-selected canonical. The selected canonical belongs to indexed information. A live test cannot predict that selection.

A third-party crawler reports what its own configuration fetched. Neither a crawler’s green result nor a successful live test proves that Google has indexed the page.

Check a canonical disagreement against Google’s canonical guidance. Canonicals are signals. Record the declared value separately from Google’s choice and fix contradictory internal links or sitemap entries before declaring the problem solved.

Trace category, variant and filter URLs

Follow a shopper from a category through pagination, product options and filtered results, recording the URL at each change. Compare the products reached by those paths with the catalogue inventory. Products present only in the inventory are orphan candidates until you find an incoming crawlable link.

For a fictional camping store, inspect the chairs category, its second results page, a folding chair and that chair’s colour options. Apply a colour filter, then combine it with a price filter. Record which URLs have distinct product sets and which reproduce the same page.

Agree the purpose of each URL before calling it defective. A useful colour category can deserve a stable search page. An empty filter combination may be deliberately excluded. A canonicalised colour variant can be intentional when the main product carries the buying information. Treat a main product accidentally pointing to an unrelated item as a fault.

Google’s ecommerce navigation guidance recommends links through category levels to products. It also explains that Googlebot generally does not submit searches into a search box. Give the developer an actual broken path and destination, then confirm that the repaired link exists in rendered output.

Audit product content and search appearance

Audit one product and one category against the shopper’s selection task, then check whether the same fault appears across their templates. Compare intended queries with useful visible information. Repeating a phrase cannot replace a missing specification.

For a fictional Shopify product targeting lightweight camping chairs, compare its title and visible heading with the product’s actual name and purpose. Confirm weight, packed size, load limit, material and stock against the catalogue record. Read the product description and search description for accurate selection details. Check identifiers, image descriptions and variant details. If a shopper cannot tell whether the chair fits their transport or use, send the missing facts to the catalogue owner.

For the matching collection, check that its title and heading describe the range it actually lists. Compare the collection’s search description with its visible range. Look for useful selection guidance and links to relevant products or subcategories. Check any collection structured data against the page’s visible information. Record an empty collection, misleading range description or inaccessible pagination separately from thin product copy.

Shopify’s SEO overview describes generated canonicals and sitemaps alongside editable product and collection information. Inspect the delivered page because themes and apps can change its output. Use the Shopify SEO guide when a confirmed issue needs platform-specific corrective steps.

Compare product structured data with visible price, availability, identifiers and selected variant. Use Google’s structured data requirements for products to choose the relevant requirements. Save a validation result and the mismatched field.

Valid markup supports eligibility for richer results. It does not guarantee their display. Assign content corrections to the catalogue owner and faulty markup generation to the developer.

Investigate rendering and mobile friction

Compare the original response with rendered output to locate where product information or links disappear. Inspect representative product, category and filtered pages on mobile as well as desktop.

A useful tech audit records the raw response, the loaded page and the rendered information available through Google’s inspection tools. If a specification is absent from the raw response but appears after JavaScript runs, investigate rendering before calling the content missing. If it remains absent in rendered output, record failed requests or script errors for the developer. Google’s JavaScript SEO documentation explains how crawling and rendering affect the content it can process.

Test the category menu, pagination, filters and variant selection on a phone-sized screen. Record overlays that cover navigation, controls that fail and layout shifts that cause wrong taps. A broken product link is a discovery defect even when the page looks complete.

Use field performance data where available and lab measurements to investigate causes. Core Web Vitals cover loading, interaction responsiveness and visual stability. An accessible but slow storefront needs performance work. A blocked storefront needs its access failure repaired first. Keep score changes separate from demonstrated crawling failures and avoid predicting rankings from a performance score.

Prioritise repairs and prove recovery

Prioritise confirmed issues by severity, commercial pages affected and repair dependencies. A shared access failure affecting the main catalogue goes ahead of optional copy improvements. Identify the template once, then list known affected URLs without presenting an untested sample as a site-wide count.

Use a report row for each distinct failure. The row needs an issue ID, affected URL and template, check, source evidence, impact, severity, dependency, owner, proposed repair and acceptance result. Keep the evidence location specific enough for another person to replay the check.

Download the SEO audit report template for Excel or another spreadsheet. The file contains the following fictional completed row. Replace its example values with your evidence. Add rows for other issues and keep a separate sheet for the sample inventory.

Issue ID,Affected URL,Template,Check,Source evidence,Impact,Severity,Dependency,Owner,Proposed repair,Acceptance result
EX-01,/collections/camping-chairs,Collection,Unintended noindex,Public response contains noindex; URL Inspection reports exclusion,Main chair collection excluded from search,Critical,Collection template release,Store developer,Remove unintended noindex from public collection template,Passed after release: noindex absent in header and rendered markup; live inspection permits indexing

The fictional acceptance result confirms a technical repair. It does not claim that rankings or sales recovered. Keep actual search recovery in a separate observation after Google revisits the pages.

After release, repeat the original failed check on the original URLs. Check neighbouring templates for unintended changes. Save the new response or rendered output beside the original evidence and give the issue a passed or failed acceptance result. Continue monitoring Search Console and organic landing-page outcomes over comparable periods.

Build an SEO audit presentation from the report in this order. Show the business symptom, bounded sample, confirmed failures, prioritised repairs and retest results. Put the evidence URL beside each finding. Label forecast revenue and vendor scores separately so they cannot be mistaken for observed losses. For ongoing category and content decisions after repairs, use the ecommerce SEO guide.

Audit internal links and backlinks as separate evidence sets. For internal links, start with broken destinations and products missing from category discovery. Record the source page, anchor, destination response and expected product. Repair the source link to the correct live destination, then recrawl that path.

For backlinks, select referring pages linked to the visibility loss or important commercial destinations. Inspect whether each page still exists, whether its link remains and whether the destination resolves correctly. Record subject relevance and the reason a lost link matters. Expand the sample when several failures share a destination or site move.

Investigate suspicious patterns through actual referring pages and any manual-action evidence. A crawler’s risk label alone does not prove that a link harms search performance. Google’s disavow guidance says most sites do not need the tool and warns that misuse can harm performance. Avoid bulk disavowal based on scores.

For a competitor audit, choose a small set of stores appearing for the same category and product tasks. Compare equivalent pages for useful selection facts, landing-page coverage and internal discovery. Record a missing load-limit comparison or absent category as a content opportunity. Another store’s unrelated article is no reason to copy its topic.

If you need a crawler to collect this evidence, compare SEO audit tools separately. Keep software rankings out of the repair sequence.

Audit international stores and a planned redesign

Audit each locale as a usable storefront, then compare its language and country signals with the corresponding pages elsewhere. Check locale URLs, visible language, local availability, canonical values and reciprocal hreflang relationships against Google’s localised-page guidance.

In a fictional store, the Australian chair page labels a New Zealand tent page as its en-NZ alternate. Correct it to the corresponding New Zealand chair page and add the reciprocal Australian reference there. Check that both destinations load and show the intended products. Investigate a canonical pointing to another locale separately before changing it.

Before a redesign, export the current URLs and useful content, including product facts, images, internal links and structured data. Map changing URLs to relevant new destinations. Document deliberate removals instead of redirecting every retired product to the homepage.

Keep staging access controlled and record the production settings that must replace staging restrictions. Google’s site-move guidance covers redirect mapping and development indexing controls. After release, request old URLs and inspect their final destinations. Check live responses, indexing directives, canonicals, sitemaps and product information.

Continue migration monitoring until Google’s observed URLs reflect the intended move. Passing launch checks alone does not establish search recovery.

Use AI audit assistance with source evidence

Use AI to group issue exports or draft explanations after collecting the source evidence. Keep confidential analytics, customer details and private URLs inside authorised tools. Supply only the fields needed for the task.

Consider this fictional false positive. An AI summary reports that a product has no canonical because its text extraction omitted the page head. The saved response contains a self-referencing canonical, rendered output preserves it and URL Inspection reports the same Google-selected canonical. Reject the finding and record that the extraction lacked the relevant evidence.

For every AI-suggested defect, return to its actual URL, response, rendered output or Search Console record. Add the issue only when that evidence supports it. Assign the repair and write its acceptance check before development starts, so the released change can be judged against the original failure.

Give your store its own shopping agent

Customers ask in their own words. Your agent finds, compares and checks out from your catalogue, right on your site.

See it in actionFree trial · Simple setup