Ecommerce SEO audit
Last updated . Figures are approximate and change with the market, so check current prices before you decide.
Stores are hard to audit because they create URLs faster than anyone can count them. Filters, sort orders, tags, variants and sold-out items each add pages. An audit separates the pages you want Google to show from the pages that only waste its time. This guide goes step by step, with Google Search Central as the main source. It works on any platform. If you want an outside team to run it, see our ecommerce SEO services.
The audit checklist
| Step | What you check | Tool | Pass means |
|---|---|---|---|
| 1. Indexing | Pages indexed and why others are not | Search Console | Money pages indexed, junk pages not |
| 2. Crawl | Status codes, redirects, titles, headings | Any site crawler | No chains, no broken internal links |
| 3. Duplicates | Canonicals and parameter URLs | Crawl plus view-source | One clean URL per page |
| 4. Facets | Filter and sort URLs | Crawl plus robots.txt | Filter URLs do not flood the crawl |
| 5. Category and product content | Text, titles, images | Manual review | Original, specific, useful |
| 6. Schema | Product structured data | Rich Results Test | Valid markup, price and availability |
| 7. Speed | LCP, INP, CLS | PageSpeed Insights | Good range at the 75th percentile |
| 8. Internal links | Paths to key pages | Crawl | Best sellers within a few clicks |
| 9. Out-of-stock | Sold-out and removed products | Manual review | Right status for each case |
Run the free website check first if you want a 30-second read on HTTPS, mobile setup, titles, headings and sitemap. It covers one page, so it is a starting point.
Step 1: Indexing in Search Console
Open the page indexing report. It shows which URLs are indexed and, for the rest, the reason. Google defines the common ones (Google):
- Crawled, currently not indexed. Google visited but chose not to index. Often thin or near-duplicate pages.
- Discovered, currently not indexed. Google knows the URL but has not crawled it yet.
- Duplicate without user-selected canonical. Google found a duplicate and picked another URL itself. Set a canonical if it picked the wrong one.
- Soft 404. A page that says "not found" but returns a success code.
Compare the indexed count with the number of products and collections you actually have. A store with 200 products and thousands of indexed URLs has a duplicate problem. A store with 200 products and 60 indexed pages has a quality or discovery problem.
Step 2: Crawl the site
Run a crawler and export the list. Look for broken internal links, redirect chains, pages with missing or repeated titles, and pages with no H1. Google asks for a unique, descriptive, brief title on each page (Google). Check that internal links are real <a href> links, because Google may not follow navigation built from clicks on other elements (Google).
Step 3: Duplicates and canonicals
A canonical tells Google which URL you prefer. Google ranks the methods: redirects are strongest, rel="canonical" next, a sitemap entry weakest (Google). Check three things on a sample of pages: the page has a self-referencing canonical, the canonical URL is absolute, and variants point to the main product. Google says variant pages can point their canonical to the main product URL, and that the URL without an optional parameter should be canonical (Google).
Step 4: Facets, filters and pagination
Faceted navigation can produce a near-endless number of URLs. Google's guidance for keeping them out of Search is robots.txt rules for the filter parameters, with a crawlable unfiltered listing page. It says canonical tags help less over time (Google). If you do want some filters indexed, such as "red running shoes", use a consistent parameter format and return a 404 for combinations with no results.
For pagination, give each page its own URL and its own canonical. Google says not to use the first page of a series as the canonical for all of them, and that it no longer uses rel prev/next (Google). Block sort orders such as ?order=price.
Step 5: Category and product content
Open ten category pages and ten product pages. Ask: does this page say anything the supplier's page does not? Google's people-first guidance asks whether a visitor would find the page useful and whether it shows first-hand knowledge (Google). Rank pages by margin or sales and rewrite the top ones first. Category pages come before single products, because they match broader searches. Check images too: descriptive filenames, specific alt text, and no important text baked into the picture (Google).
Step 6: Structured data
Google's ecommerce docs point to Product markup, with ProductGroup for variants (Google). Test a product, a category and the home page in the Rich Results Test. Look for missing price, availability or image. Then compare the markup with what the visible page shows, since they should match.
Step 7: Speed and Core Web Vitals
The targets are an LCP of 2.5 seconds or less, an INP of 200 milliseconds or less and a CLS of 0.1 or less, at the 75th percentile, split by mobile and desktop (web.dev). Common causes of slow LCP include a slow server, an image that loads late, large files and render-blocking scripts. web.dev warns never to lazy-load the LCP image (web.dev). On stores, the hero image and third-party apps are the usual suspects.
Step 8: Internal links
Google reads links between pages to judge importance. It says the number of internal links to a page and how many clicks it takes to reach it are signals, and suggests linking best sellers from the home page or content such as blog posts (Google). In your crawl, find product pages with no internal links, and key collections that sit deep in the menu. Products reachable only through the site search may never be found.
Step 9: Out-of-stock and removed products
Treat each case differently:
- Back soon. Keep the page live, say when it returns, and mark availability in the markup. Google's John Mueller has described this as a valid approach (Search Engine Journal).
- Gone for good, with a close replacement. Redirect to the replacement.
- Gone, nothing similar. Return a 404.
- Empty category. Google says to use
noindexfor an empty category, or a 404 if you removed it (Google).
Turn findings into a fix list
Score each finding by pages affected and effort. A canonical fix that touches 2,000 URLs beats a title rewrite on one. Put the quick wins first, then the content work. Re-check Search Console after each batch, and expect crawling changes to take days to weeks.
FAQ
How often should I audit an ecommerce site?
Do a full pass once a year and after any platform, theme or URL change. Check Search Console monthly for new indexing errors.
What do I need for an ecommerce SEO audit?
Search Console access, a crawler, PageSpeed Insights and the Rich Results Test. The first and the last two are free.
What is the most common problem on stores?
Often, too many low-value URLs from filters, tags and variants, and category pages with almost no text. Your Search Console report will show whether that applies to you.
Can I audit a Shopify store the same way?
Yes. Shopify adds platform-specific checks, which we cover in our Shopify SEO checklist.
Should I fix everything the audit finds?
No. Fix what affects many pages or key revenue pages first. Leave small issues that cost more to fix than they return.
How OnePDA helps
Our ecommerce SEO services begin with this audit: indexing, speed, duplicate pages from filters, pagination, redirects and errors, ranked by how much each matters. Pricing is from €490 a month with a 3-month minimum, and guide articles are €59 each. A monthly report shows clicks, impressions, positions and orders from search. We do not promise rankings, and the accounts stay in your name. Prices exclude VAT where it applies.