Best SEO audit tools for ecommerce 2026

Choose the best SEO audit tool for your store. Compare catalogue crawl coverage, rendering, issue evidence and the limits of free software.

Screaming Frog SEO Spider is the best SEO audit tool for ecommerce catalogues because you can trace a product-page issue back to its source. Its paid edition combines JavaScript rendering with detailed URL exports. You can inspect canonical tags and the links pointing to a product, then send your developer the evidence behind a fix.

A hosted audit suite makes recurring checks easier to share across a team. A guided crawler helps when you need explanations beside the warnings. The choices below suit different catalogue workflows, from a small HTML store to product pages that depend on JavaScript.

Top SEO audit tools for online stores

1. Screaming Frog SEO Spider

For detailed catalogue investigations, Screaming Frog SEO Spider gives you direct control over the evidence you collect. It runs on your computer and lets you control what it crawls. For a colour variant with the wrong canonical, you can inspect the declared target and find the pages linking to that variant. That gives a developer a URL-level starting point instead of a health score alone.

The licensed edition adds rendered crawling, structured-data validation and custom extraction. Extraction is useful when you need product identifiers beside price or availability. You can group exported rows by a product-template field you collect, then investigate a shared defect once. Choosing the right fields takes technical knowledge, and your computer’s memory and storage limit large crawls.

The free edition crawls up to 500 URLs per crawl, without an account signup. It checks basic response and metadata issues, but rendering and saved crawls require a licence. The paid edition also supports scheduling and crawl comparison. Choose it when a technical member of your team needs repeatable evidence and control over the crawl, rather than a hosted reporting workspace.

2. Sitebulb

Sitebulb ranks second for teams that need help interpreting a technical finding. Its Hints explain issues beside the affected URLs. The Response vs Render report separates elements created or changed by JavaScript, so you can investigate a canonical that changes after the page loads.

Its structured-data reports distinguish Schema.org validation from Google search-feature requirements. URL lists and exports help you collect affected product pages, while its page-template analysis helps separate repeated warnings from their shared cause. This suits a catalogue manager reviewing a developer’s findings. The explanations still need your judgement about which products and filters are meant to appear in search.

Sitebulb Desktop has a 14-day Pro-equivalent trial that expires without automatically becoming a paid subscription. Desktop Lite includes JavaScript crawling with a 10,000-URL audit limit. Pro lists 500,000 URLs per audit and adds audit comparison and scheduling. Cloud runs remote crawls and supports team collaboration, with its own subscription. Choose the edition for where the crawl must run, as well as the report you want to read.

3. Ahrefs Site Audit

Ahrefs Site Audit is third because its continuing free access makes scheduled checks practical for a retailer’s own verified store. Ahrefs Free includes 5,000 crawl credits per month for each verified project and requires no credit card. That can cover a small catalogue repeatedly when the successful HTML-page count stays within the allowance.

The hosted crawler has JavaScript rendering and page-level reports. Use issue details to find a suspect product, then inspect its canonical and inlinks in the URL details. Page Explorer filters help isolate a product directory. This is useful for watching a recurring template issue without keeping a desktop crawl running.

Crawl credits count internal HTML pages returning HTTP 200. Redirects, error responses and resources do not consume those credits. The free allowance is a monthly budget, so repeating the same successful pages spends it again. Exports need separate attention: an overview PDF summarises the crawl, while affected-page exports carry the repair evidence. The account’s export permissions decide what you can take out, so free crawl capacity alone does not establish a complete handoff workflow.

4. Semrush Site Audit

Semrush Site Audit ranks fourth for a retailer that wants technical checks beside a wider SEO suite. Its crawl configuration accepts a defined URL file, which suits a repeatable catalogue sample. Issue reports connect warnings to affected pages, and the export choices include page-level issue columns.

Its JS Impact report makes changes after rendering easier to inspect. A product template that changes its canonical or removes internal links deserves a closer look even when the final page appears normal. JavaScript rendering requires the Guru or Business SEO Toolkit tier. Pro’s larger crawl allowance does not include that rendering entitlement.

The suite is a practical choice when your team already uses its other reports. Buying it solely to inspect a JavaScript product template requires a comparison with a dedicated crawler’s paid edition. A free account has a 100-page monthly crawl allowance. That is enough for the example below once, but too small to infer coverage of a substantial store.

The ranked tools differ in crawl scope and the evidence they return.

ToolCrawl scope and renderingProduct schema and evidenceIssue handoffFree access and paid limit
Screaming Frog SEO SpiderDesktop site or URL-list crawl; rendering requires a licenceLicensed structured-data checks and custom extractionURL exports, source details and inlinksFree: 500 URLs per crawl; licensed capacity depends on computer resources
SitebulbDesktop or Cloud; JavaScript crawler availableSchema and Google search-feature checks; Response vs RenderHint and URL-list exports, including CSVDesktop trial: 14 days; Lite: 10,000 URLs per audit; Pro lists 500,000
Ahrefs Site AuditHosted project; rendering settingStructured-data checks, issue pages and URL detailsOverview PDF and affected-page exports, subject to account accessFree verified project: 5,000 credits monthly; Lite lists 100,000 monthly credits and 25,000 pages per project
Semrush Site AuditDomain, folder, sitemap or imported URLs; rendering on Guru/BusinessMarkup reports and JS Impact evidenceCrawled Pages with Issues file, CSV/Excel and PDFFree: 100 pages monthly; Pro/Guru: 20,000 per audit; Business: 100,000

For an Australian store, crawl the Australian storefront and keep its locale consistent. A region redirect or a currency cookie can change what the crawler receives. Compare the displayed price currency with the product offer markup, and use Australian country filters in connected Search Console performance data.

An Australian keyword database does not determine whether a crawler can inspect your robots directives or declared canonicals. Local search-demand research is a separate job. Compare support availability against your working hours, plus the subscription currency and any tax or extra-seat charges, before choosing a hosted suite.

Trial each tool on variants, filters and product templates

Use a bounded catalogue sample so every tool answers the same question. The following fictional sample contains ten URLs. Its paths illustrate store behaviour, not crawl results from a real retailer.

URL pathCatalogue roleEvidence to inspect
/collections/coats/Main collectionLinks to products and the next collection page
/collections/coats/?page=2PaginationReachable links to later products
/collections/coats/?colour=navyColour filterIntended indexing policy and canonical target
/collections/coats/?colour=navy&size=mCombined filtersIntentional exclusion versus an accidental block
/products/rain-coat/Product parentCanonical and ProductGroup relationship
/products/rain-coat/?variant=navy-sSmall navy variantVariant identity, selected size and offer data
/products/rain-coat/?variant=navy-mMedium navy variantIts own availability and selected option
/products/rain-coat/?variant=red-mDifferent colourColour-specific product data and imagery
/products/old-parka/Unavailable productResponse, availability and links to purchasable alternatives
/products/packable-jacket/JavaScript-dependent templateProduct content and links before and after rendering

Start with link discovery from the collection, then compare its discovered URLs with the ten-URL list. An imported URL can be inspected even when no crawlable link leads to it. That difference exposes broken discovery without confusing it with a failed page fetch.

For each URL, retain the response status and robots evidence alongside the canonical. Record the parent/variant schema relationship, visible price and availability, then inspect links to purchasable products. Google’s variant structured-data guidance explains how ProductGroup connects variants with their parent.

Suppose the fictional medium navy variant declares an unrelated coat as its canonical. The useful evidence is the affected URL and the exact canonical target, with the selected variant and template group. A generic duplicate-page count cannot explain why the wrong item could become the preferred search destination.

A filtered collection deliberately excluded from indexing can be correct. An excluded product that your team intends to sell through search needs investigation. Agree those intentions before treating every warning as a defect. Google’s faceted-navigation guidance explains why uncontrolled filter combinations can consume crawl resources.

Screaming Frog’s extraction fields can carry a template identifier into an export. Sitebulb’s template reports and Ahrefs’ directory filters help inspect repeated patterns. In Semrush, filter affected-page rows by product path and join them to your catalogue’s template mapping.

Sitebulb’s template grouping excludes query-string and paginated URLs, so map those sample rows to templates using your catalogue data. These approaches help distinguish one faulty template from many separate repairs. Findings from ten URLs still describe only those ten URLs.

In Screaming Frog List mode, record robots handling explicitly: its documented default ignores robots.txt for uploaded lists. Use the licensed configuration to preserve the intended baseline. An allowed inspection of a blocked URL must remain distinct from ordinary crawler discovery.

Evaluate Semrush site audit on the same store sample

Use a Semrush account with enough remaining crawl allowance for the sample and its repeat run. Guru or Business SEO Toolkit access is needed for rendered inspection. The following setup uses the fictional ten-URL catalogue and illustrative evidence.

  1. Open Site Audit and create or select the store project. Set Scope to the Australian storefront domain or relevant subfolder. Clear all-subdomain crawling when those hosts are outside the sample.
  2. Set Limit per audit to cover the ten sample URLs. Under Pages to crawl, choose Import URLs from file and upload a CSV or TXT file containing one full URL per line. Confirm the detected URL count.
  3. Retain the sample’s variant, colour, size and pagination parameters. Parameters to ignore strips the named parameter from crawl URLs, so ignoring these would collapse cases you need to compare.
  4. Set the mobile crawler and a crawl delay suitable for the store. Keep robots bypass disabled for the baseline. Use allow/disallow rules only where they preserve the agreed sample.
  5. Run the HTML crawl first. On an eligible subscription, enable JS rendering for a second run of the same scope. Inspect JS Impact for changed canonicals and internal links.

Semrush’s configuration guide documents these controls. A website-source crawl answers discovery questions separately from the imported-list inspection. Keep the collection as its starting point and restrict expansion to the agreed paths so filters do not create an uncontrolled crawl.

For the fictional wrong-canonical finding, open the relevant issue and its affected pages, then inspect the variant’s page details. Retain the canonical target shown and the crawl’s rendering mode. If the target changes after JavaScript runs, JS Impact supports investigating that difference. It does not establish Google’s chosen canonical.

Export Crawled Pages with Issues to attach affected URLs and issue fields to the repair ticket. Add the relevant markup export when the finding concerns product schema. Semrush’s export documentation distinguishes those files from a presentation PDF. After the fix, rerun the same scope and use Compare Crawls to inspect the changed issue.

An HTML-only tier can inspect a canonical already present in the response. It cannot establish whether this product template changes that canonical after rendering. Semrush suits a team that wants recurring audit reports beside its wider SEO work. A dedicated crawler suits the narrower job when detailed extraction and local control decide the purchase.

Compare Semrush, WooRank, Moz, Raven, Ahrefs and Yoast

This unranked matrix covers the named products’ different jobs. A site review, an editor plugin and a catalogue crawler do not provide equivalent evidence.

ProductScope and JavaScriptCatalogue evidenceAccount and usageExport method
Semrush Site AuditProject scope or imported URLs; Guru/Business renderingAffected pages, canonical and markup reports, JS ImpactFree 100 pages monthly; Pro 100,000 and Guru 300,000 monthly; per-audit caps also applyTop-right Export for page/issue files; PDF button for report views
WooRank Site CrawlProject crawl; use rendered checks separately unless the selected product demonstrates themOn-page fields, status, indexability and canonical issuesProject access with Site Crawl entitlement; crawl and project capacity depend on the subscriptionDownload completed Site Crawl results from the crawl view
Moz Pro Site CrawlCampaign crawl follows HTML links; JavaScript-only links can be missedCanonical target, redirects, robots fields and duplicate-content groupStandard lists three tracked sites and 100,000 pages weekly; monthly headline is 400,000Site Crawl, All Crawled Pages, Export CSV or PDF
Raven Site AuditorWebsite crawl with path exclusions; documented crawler does not follow JavaScript linksVisibility, metadata, content, link, image and semantic issuesHelp lists 10,000 pages per website; Start has 400,000 pages per billing cycleStudio shareable report; reporting also supports HTML/PDF handoff
Ahrefs Site AuditHosted project with rendering optionIssue details, Page Explorer and individual URL evidenceVerified projects have 5,000 free credits monthly; paid project/page caps differPrint overview to PDF; open an issue’s affected pages and select Export where entitled
Yoast productsPublic single-page checker or installed platform product; editor analysis is distinct from rendered crawlingPage checks, metadata controls and platform-specific schemaWordPress installation, WooCommerce extension or Shopify app subscription, according to productPlatform Import and Export for SEO data/settings; these are not crawl-issue exports

WooRank separates its broad site review from Site Crawl’s URL-level investigation. Its help describes canonical and indexability issues, including deliberately blocked pages. For the fictional variant, inspect the affected URL and its canonical evidence. Use the completed crawl’s download for handoff, then refresh the crawl after repair. The documented weekly or fortnightly scheduling and manual refresh support repeat checks.

A “Google Vision” label does not establish JavaScript rendering. For a variant whose price appears only after scripts run, pair WooRank’s crawl with rendered inspection unless your selected product demonstrates that content. Do not substitute a review PDF or keyword-report CSV for the crawl’s affected-page evidence.

Moz Pro suits source-HTML investigations inside a tracked campaign. Its All Crawled Pages CSV includes the referring URL and status, plus canonical target, redirect location and robots fields. Duplicate-content group identifiers help investigate related product rows. These fields can reveal the fictional wrong canonical, while a JavaScript-only link can escape discovery. Separate Lighthouse performance reports do not remove that crawler limitation.

Moz’s pricing distinguishes weekly page capacity from the monthly headline. Standard has no Site Re-crawls allowance in its comparison table, while Medium lists ten. Export the relevant CSV before the next campaign crawl: Moz’s help says a new crawl replaces the previous detailed data. Historical graph counts remain available, but the old URL inventory cannot then be exported.

Raven Tools supports catalogue issue handoff through its audit reports. Its Classic and Studio views present the audit differently, and Studio supports a shareable report. A visibility issue can identify a broken product URL, while content or semantic categories help investigate repeated product-page problems. Retain the affected-page detail with the sample’s product-template mapping.

Raven’s published monthly allowance is separate from the per-website cap in its help. A 400,000-page billing-cycle budget does not turn a 10,000-page website crawl into a 400,000-page crawl. Path exclusions can keep unrelated store areas out. Its help explicitly limits discovery to ordinary hyperlinks, so JavaScript-only variant links need another inspection route.

Ahrefs distinguishes the overview PDF from exports opened inside an individual issue’s affected-page report. Use the latter for a wrong-canonical ticket. Its product page also advertises bulk issue downloads, while the export help describes separate issue exports. For a repeatable handoff, use the documented issue-level route and test that output on the account you will operate.

The Ahrefs Free and Starter comparison lists zero general export rows. That table alone does not settle every Site Audit output permission. Keep the PDF-versus-URL-export distinction when evaluating free access, and choose a paid tier only when its actual export entitlement meets the handoff. Do not buy a general report-row allowance on the assumption that it measures crawl credits.

Yoast SEO has a public checker for a single submitted page. It checks on-page elements and technical signals such as canonical tags. The free WordPress plugin adds editor analysis and page controls. Its JavaScript analysis library analyses content in the editor. That does not make it a browser-rendering catalogue crawler.

Yoast WooCommerce SEO adds product-specific analysis and structured-data support to the WordPress store workflow. Yoast SEO for Shopify runs in the Shopify admin and adds product schema and content guidance through a subscription. These products help change the template or product data behind a finding. Their metadata or settings exports do not replace a crawl report listing affected URLs.

For the fictional navy variant, inspect the relevant product’s editor data and page controls, then compare the public page’s canonical and offer data. Keep its template mapping beside that evidence so a fix reaches other affected products.

See what crawl reports leave out

A crawl records what that crawler fetched under its settings. Raw-response crawling inspects returned HTML. Rendered crawling executes scripts before inspecting the resulting document. Blocked script resources or a short wait can leave product content absent even in a rendering crawl.

Rendering also differs from a customer clicking a filter or selecting a size. A crawler that loads the initial page does not necessarily exercise those controls. Inspect the state-specific URLs and compare the selected variant with its offer data. Google’s JavaScript SEO guidance explains the separate crawling and rendering stages.

A declared canonical is the store’s signal. Search Console’s URL Inspection report shows Google’s selected canonical in indexed data. Its live test cannot predict that selection. Server or content delivery network logs establish actual requests and response failures at the store, after validating the crawler identity.

A store crawl’s external links show where the store links out. Evaluating referring sites or lost backlinks needs an external backlink index, such as the vendor’s separate backlink product. A Search Console integration can add search-performance evidence, but it does not turn every crawl observation into Google indexing data.

A clean report or valid schema does not prove full indexation or guarantee rich-result display. A health score cannot establish a ranking gain. Product stock also needs comparison with the retailer’s stock source: a page can consistently publish incorrect availability in both visible text and schema.

Compare free access with paid crawl capacity

Ahrefs Free is continuing verified-site access. Screaming Frog’s free desktop edition is restricted by features and per-crawl size. Sitebulb’s trial provides temporary Pro access, while a paid subscription keeps those features available. Treat these as different ways to complete a job.

For the fictional ten-URL sample, free HTML crawling can inspect status and canonical signals. It cannot complete the rendered-product comparison in Screaming Frog’s free edition or Semrush’s free account. Ahrefs’ verified-project allowance can cover the sample’s successful HTML pages, and Sitebulb’s Pro-equivalent trial can cover its rendered inspection.

Plan capacity from crawled URLs and repeat runs. In an illustrative store, 200 product parents with four URL-addressable variants each already create 1,000 product URLs before collections and filters. Two crawls repeat that workload. Ahrefs charges credits only for qualifying successful HTML pages, while other products use their own page or URL allowances. Keep those units separate.

Screaming Frog lists USD $279 per user for a one-year licence, with no monthly crawl-credit budget. Tax treatment depends on checkout. Sitebulb Desktop subscriptions separate Lite from Pro, and Cloud has its own plan. Count extra seats and the machine or Cloud resources needed to run the intended crawl.

For a hosted suite comparison, Moz Standard lists USD $99 on monthly billing and three tracked sites. Its page capacity is 100,000 weekly, displayed as 400,000 monthly. The displayed rate does not state a universal Australian tax total. This pays for a wider suite, so compare the work your team will use.

Yoast SEO for Shopify lists USD $19 per 30 days excluding VAT, with a 14-day trial. It is a platform SEO app rather than equivalent crawl capacity.

Budget the foreign-currency charge and any applicable checkout tax separately from the advertised rate. A prepaid annual equivalent is not a month-to-month subscription. Optional paid data or extra users can increase a suite’s bill. Ahrefs documents automatic charges when enabled pay-as-you-go usage exceeds the plan.

Saved crawls and scheduling affect operating cost too. A free report that cannot preserve the evidence you need can create manual work at every release. For Semrush’s wider subscription allowances and costs, use the Semrush pricing guide after establishing which audit tier you need.

Check whether issue evidence is useful to your team

Choose an SEO audit report tool by whether its exported finding supports a repair. For the fictional canonical error, keep the affected variant and canonical target together. Include the source or rendered evidence, rendering mode and a collection link that points to the variant. Add the template identifier so the developer can inspect other pages using it.

A concise illustrative repair ticket can contain the following fields.

Ticket fieldIllustrative handoff
Affected templateCoat variant product template
Example URL/products/rain-coat/?variant=navy-m
Observed evidenceA hypothetical crawl row declares the old parka as canonical; attach the corresponding HTML evidence
Expected behaviourCanonical follows the store’s agreed parent/variant policy and never selects the unrelated parka
Acceptance checkRepeat the ten-URL crawl in the same rendering mode; confirm the intended target and inspect the selected variant
Customer impactThe intended coat variant’s search signals point to an unrelated item

This ticket illustrates the evidence required. It is not an observed vendor finding. Add a separate ticket when the unavailable parka’s page claims stock that the retailer cannot fulfil. That issue has a different customer consequence from an intentional noindex on a combined filter.

Preserve the original URL list and crawl settings with the export. Changing the page allowance can reduce warnings by omitting affected URLs. Turning rendering off can hide a JavaScript-created defect. Compare the same population and inspect the repaired examples before treating a lower issue count as progress.

The broader ecommerce SEO audit guide covers the diagnostic and repair sequence after software selection. Use this trial to judge the quality of the evidence you can pass to that process.

Match the subscription to your catalogue

A small stable range can start with a bounded free crawl when its product content is present in HTML. Include every template and an unavailable product, then confirm the report can preserve the findings you need. A successful small sample justifies expanding coverage, not claiming the whole store is clear.

A variant-heavy store benefits from a licensed dedicated crawler when parameter control and rendered evidence decide the investigation. Screaming Frog suits detailed extraction and technical ownership. Sitebulb suits a team that needs guided explanations and template-level investigation. Use the trial or licence on the catalogue sample before committing to full-store capacity.

A frequently changing catalogue benefits from scheduled hosted checks and a repeatable handoff. Ahrefs suits ongoing verified-store monitoring within its credit budget. Semrush suits a team already using the wider suite, with Guru or Business needed for rendered product pages. Preserve sample settings between releases so the comparison remains meaningful.

Keep Search Console and request logs available for questions the chosen crawler cannot settle. Researching product or category demand needs a different dataset, covered in the guide to keyword research tools. Start with the relevant tier on your representative URLs, export one repairable finding and repeat the crawl after its fix.

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