Google normally consolidates duplicate or very similar URLs instead of punishing them. When it finds several pages that appear to be the same, it groups them and chooses one representative canonical URL (Google Search Central, how Google canonicalizes URLs).
Google also says that ordinary duplicate content is normal, is not a spam-policy violation, and does not cause a manual action. Its concern is efficiency, not a duplicate-content penalty (Google Search Central, SEO starter guide). Nobody should sell you penalty recovery for accidental store duplication.
What it does cost an ecommerce store is four specific things:
- Unnecessary crawling. Google recommends reducing alternate ecommerce URLs that return the same content so its crawler does not make more requests than needed (Google Search Central, ecommerce URL structure).
- Split or contradictory signals. A declared canonical that disagrees with your internal links, redirects, and sitemap is a preference nobody can read cleanly.
- Harder reporting. Impressions and positions scatter across URL forms, so it becomes difficult to tell whether a page is underperforming or merely being counted somewhere else.
- A representative URL you did not choose. A declared canonical is a preference rather than a binding instruction. Google can select a different URL from the group, and Search Console reports that case.
So the useful question is not "do we have duplicate content." Every catalog does. It is which duplicate URLs are consuming crawling, contradicting each other, or being represented by the wrong page.
Eight ways an ecommerce store generates it
Nearly all store duplication traces back to one of these. Several are normal platform behaviour; the evidence in the triage table determines whether any change is justified.
- Product variants. Size, colour, and option selections that each produce their own address. Google supports variant URLs and documents both path segments such as
/t-shirt/greenand query parameters such as/t-shirt?color=green. For optional variant parameters, it recommends the parameterless URL as the canonical. - Collection-path product URLs. The same product reachable through several category or collection paths, so one item has two or more addresses that render the same page.
- Filter, sort, session, tracking, and search parameters. Faceted navigation, sort orders, session identifiers, and campaign tags multiply URLs quickly. Google warns that continually changing URL values can create what appears to a crawler to be an unbounded number of pages.
- Near-identical categories, tags, and collections. Two taxonomy pages listing broadly the same products for broadly the same shopper. This is a merchandising decision that presents as a duplication symptom.
- Supplier or manufacturer descriptions. The same product copy carried by many retailers. This is a cross-domain condition, not a same-site URL condition, and it behaves differently. See the section below.
- Repeated template titles and meta descriptions. One title or description pattern applied across a catalog, so similar pages become indistinguishable in results. Google recommends page-specific titles and descriptions.
- Print views, paginated pages, and internal search results. Alternate renderings and generated listing pages that expose the same inventory under different addresses.
- Cross-domain feed and marketplace copies. The same listing published to a marketplace, a comparison feed, or a partner site. Your store does not control those URLs, and the decision there is different from anything on your own domain.
Triage table: what to collect before deciding anything
Read the fourth column as the honest default. "Usually resolves" does not mean ignore it; it means the evidence has to show a problem before a change is justified.
| Source | How it appears | Evidence to collect | Consolidation usually expected | Next decision |
|---|---|---|---|---|
| Product variants | Query strings or path segments on a product URL, one per option combination | Which variant forms are crawlable, what each declares as canonical, and whether internal links use the parameterless form | Yes, where the declared canonical and the internal links agree | If variants have genuinely distinct demand, keep them as real URLs. If they are one buying decision on one page, point them at the parameterless product URL and link that way |
| Collection-path product URLs | One product reachable at two or more category paths | Which form collection templates link, what the product page declares, and which form appears in the sitemap | Yes, when one form is declared consistently | Pick one form and use it consistently in navigation and the sitemap. Re-check only if Search Console or a later template change shows disagreement |
| Filter, sort, session, and tracking parameters | Query strings appended by facets, sorting, sessions, or campaigns | How many parameter forms are crawlable, whether they are internally linked or only discoverable by crawling, and whether any are indexed | Partly. Crawling still happens even when indexing consolidates | Decide which facet combinations answer real demand before blocking anything. For most stores this is a crawl-efficiency and navigation decision more than a duplication one |
| Near-identical categories, tags, collections | Two taxonomy pages with interchangeable titles and largely the same products | Product overlap between the two, the titles and H1s, the descriptive copy on each, and which one receives internal links | No. These are distinct pages that are simply too similar to each other | Merge, differentiate, or set a deliberate parent and child. A canonical tag is not the answer here |
| Supplier or manufacturer descriptions | The same paragraphs on your page and on many other retailers | An exact-phrase search on a distinctive sentence, plus what your page adds that the others do not | Not applicable. Nothing on your site is duplicated | Treat as a differentiation question, not a canonicalization one. Judge whether the page gives a shopper a reason to buy here |
| Repeated template titles and descriptions | Dozens of pages sharing one title or description pattern | The rendered title and meta description across several pages of the same template | No. The URLs are distinct; the metadata is not | A template edit, not a URL change. Give the template enough page-specific input to produce a distinct title |
| Print views, pagination, internal search | Alternate rendering paths and generated listing URLs | Status codes, whether they are linked, whether deep products are reachable only through pagination, and whether search URLs are indexed | Mixed. Pagination is not duplication; print views and search URLs often are | Make sure deep inventory remains reachable through ordinary links. Decide separately whether search and print URLs should be crawlable at all |
| Cross-domain feed and marketplace copies | The same listing on a marketplace, feed, or partner site | Where the copies are, which one search results currently show, and whether your own URL is indexed at all | Not on your terms. You do not control the other domain | Confirm your own page is indexed and useful first. Distribution choices about marketplaces are a business decision, not a fix |
The pattern: rows one, two, three, and seven are usually handled by making your own signals agree with each other. Rows four and six need editorial or merchandising work. Rows five and eight are not duplicate-URL problems at all, and treating them as such is where most wasted effort goes.
What a canonical settles, and what it does not
Google treats redirects and rel="canonical" links as strong canonicalization signals and sitemap inclusion as a weaker one. Consistent signals reinforce each other (Google Search Central, specify a canonical with rel=canonical and other methods).
Strong is not the same as binding. Four things follow from that, and they are what the tag is repeatedly asked to do and cannot.
- It states a preference, not an instruction. Google documents canonical selection as a hint rather than a rule.
- It works best when the rest of the site agrees. Google recommends using the same preferred URL across canonical links, redirects, internal links, and sitemaps. Contradictory signals make that preference less clear and can lead Google to select a different URL.
- It is not a crawl control. Duplicate URLs may be crawled less often, but canonicalization does not stop crawling. Google also advises against using robots.txt for canonicalization or pointing canonicals at noindexed or redirected pages.
- It does not make two similar pages worth having. Choosing between two near-identical collection pages is not the same as making either of them useful. That work is editorial.
The procedural version of this check, including what to record for each sampled URL, is step 4 of the free checklist: test canonicals, filters, variants, and parameters. This guide diagnoses the condition; that step runs the procedure. If the duplication started at a relaunch rather than accumulating, the redesign and migration guide covers the launch-day canonical and redirect validation instead.
Supplier and manufacturer descriptions
This is the case most often mislabelled, so it is worth separating cleanly.
Same-site duplicate URLs
Two or more addresses on your domain render the same or very similar content. Google clusters them and picks one to represent the group. The lever you hold is signal consistency: redirects, canonical links, internal links, and sitemap membership all pointing the same way.
Cross-domain repeated copy
Your page carries text that also appears on many other retailers because the manufacturer supplied it. Nothing on your site is duplicated. Google is choosing between competing pages on different domains, and the question is which page best answers the search, not which URL represents a cluster.
Google's starter guide separates ordinary same-site duplicates from copied third-party content. Supplier copy sits between those cases: it is generally provided for retailers to use, so it is not necessarily scraping, but the page may still offer little that a shopper cannot read elsewhere.
When supplier copy is likely to be the actual problem
- The product page is indexed and reachable, but is not ranking for the product name against other retailers carrying the same listing.
- An exact-phrase search on a distinctive sentence from your description returns a long list of stores, and yours is not among the first results.
- The page contains the supplier paragraphs and effectively nothing else: no sizing or compatibility detail, no comparison against the alternatives you also stock, no shipping or returns specifics, no images or reviews of your own.
- Your category and collection pages carry the same problem, because their copy was also written once and applied everywhere.
When it is probably not the problem
- The page already carries substantial store-specific detail alongside the supplier text.
- The product has almost no search demand under any phrasing, in which case rewriting the copy changes nothing.
- The page is not indexed at all, which is an indexability question that has to be settled first.
- You are the only retailer carrying the item, so there is no competing copy to be distinguished from.
Scope boundary, stated plainly. StoreAuditLab does not sell product-copy writing, description rewriting, or content production of any kind. The audit can tell you whether repeated supplier text is a plausible cause on the pages it reviews, and what a differentiated version of that page would need to contain. Writing it is your side of the line.
How to verify a small sample yourself
Six checks over a handful of representative URLs. Pick one variant-heavy product, one product reachable through more than one collection, two collection or category pages that feel similar, one filtered or sorted URL, and one paginated URL. That is enough to expose template behaviour without pretending to be a crawl.
Read the Page indexing report
Search Console's Page indexing report names three useful conditions: Alternate page with proper canonical tag, Duplicate without user-selected canonical, and Duplicate, Google chose different canonical than user. The first is often normal consolidation. The other two are prompts to inspect the affected examples, not automatic proof that a change is required.
Compare the two canonicals in URL Inspection
The URL Inspection tool reports both the User-declared canonical and the Google-selected canonical. A difference is worth investigating against the page's redirects, links, and sitemap. A match is evidence that the sampled URL is consolidating as intended.
Compare rendered source, not the browser view
View the rendered HTML of each sampled URL and record the title, the H1, the canonical href, and the robots directive. Two URLs that look identical on screen can declare different canonicals, and a variant URL that looks like a duplicate may declare the parameterless product URL correctly. The declaration is the evidence; the visual is not.
Check which form your internal links use
Follow a product from the collection grid, the navigation, and any guide that mentions it. If the three arrive at different URL forms, your site is voting against its own canonical. Google recommends using the same preferred form in internal links, sitemap files, and canonical tags.
Check sitemap membership and status codes
Confirm the sitemap lists the URLs you deliberately want indexed and excludes redirects or alternate parameter forms that are meant to consolidate elsewhere. Then confirm each sampled URL returns the status code you expect. A canonical pointing at a redirect, or a sitemap listing a URL the site otherwise treats as an alternate, is a contradiction you can find in minutes.
Read the sample as a sample
Six or ten URLs can establish that a template behaves a certain way. They cannot establish how many URLs on the site are affected, which of them Google has crawled, or what the total crawlable surface is. Treat a clean sample as "no template-level problem found here," not as "the store has no duplication."
Email me the duplicate-content diagnostic worksheet
Use the two-page printable PDF to compare declared canonicals, Google-selected canonicals, internal-link forms, sitemap membership, and status codes across a small URL sample.
Short platform notes
The evidence-led process above does not change by platform. What changes is which of the eight sources your platform produces by default.
Shopify
Shopify's SEO overview says the platform generates canonical tags, sitemap.xml, and robots.txt automatically. Its robots.txt documentation describes default rules that block several filtered collection, search, cart, and checkout paths. Defaults are not verification: the audit still reads what a given theme and its apps emit on the live page. More on the platform in the Shopify URL checks.
WooCommerce
WooCommerce and WordPress together produce more surfaces than either does alone, and the output depends on the permalink choice, the theme, and whichever SEO plugin is installed. That means there is no single default worth quoting here. Inspect the product permalink base, category, tag and attribute archives, filtered and faceted URLs, and how variable products expose their variations. The WooCommerce archive checks and variation checks cover those surfaces in context.
BigCommerce is the third platform the audit covers, and it belongs in the same category: a crawlable storefront where the same evidence, collected the same way, answers the same questions. Nothing in this guide is platform-specific except which defaults you start from.
When duplication is not the problem
This is the section that decides whether the work is worth doing. Five common cases where duplicate URLs are visible and are not what is holding the pages back.
| What you are seeing | Why it is probably not duplication | What to look at instead |
|---|---|---|
| Search Console lists many pages as alternates with a proper canonical | Google describes that label as a page correctly pointing to an indexed canonical | Nothing. This is consolidation working. Volume in this label is not a defect count |
| Variant URLs are indexed separately | Google supports variant URLs and documents both a path-segment and a query-parameter form | Whether shoppers actually search for the variant. Genuinely distinct variants deserve distinct URLs |
| A product page is indexed but gets no impressions | A page that is consolidated correctly and still invisible has a different constraint | Whether the page says anything a shopper needs, and whether the product has demand under any phrasing |
| Collections exist but nothing links to them | Orphaned pages fail on discovery and importance, not on similarity | Navigation paths and internal links from the guides that should feed each collection |
| Everything is technically clean and traffic is still flat | Consolidation does not create demand | Whether search demand exists for the terms the pages target, and whether the page type matches the intent behind them |
There is also no documented percentage to hit in Google's current public guidance. Treat any specific number as an estimate, not an official limit. The workable test is whether a given URL is worth crawling, worth indexing, and worth a shopper's time. That question does not have a percentage answer.
What a review of public pages can establish here
StoreAuditLab sells one fixed-scope product: a manual review of up to 10 representative public URLs, plus up to 10 Ahrefs-backed keyword and page opportunities, delivered as a prioritized PDF within 48 hours for $100. On this topic specifically, that scope has a clear edge.
Observable from public pages
- What each sampled URL declares as its canonical, and whether the declarations are consistent across a template
- Whether internal links, navigation, and the sitemap point at the same URL form the canonical names
- Which variant, filter, and collection-path forms are reachable and what they return
- Whether two collection pages carry interchangeable titles, descriptions, and inventory
- Whether a template is writing the same title or description across many pages
- Whether product copy appears to be supplier text carried by other retailers
Requires access we do not have
- Which URL Google actually selected as canonical, which only Search Console reports
- How many URLs on the site are affected, which needs a complete crawl
- What Google has crawled and how often, which needs Crawl Stats or server logs
- Whether duplication caused a traffic change, which needs Search Console history and analytics
- Any ranking, traffic, or revenue outcome from making a change
The practical split: a public-page review is good at finding the contradiction and naming the template it lives in. Confirming what Google did with those signals is your Search Console, and it is free. What sets an audit's price covers the rest of the scope boundary, and the full sample report shows how an indexation and canonical finding is actually written up, in section 4 of the report.
Common questions
Is there a duplicate content penalty?
No. Google says ordinary duplicate content on a site is normal, is not a spam-policy violation, and does not cause a manual action. The practical concern is efficiency and clarity: alternate URLs can consume unnecessary crawling or send inconsistent canonical signals. Copying another site's content is a separate issue.
Should product variants have their own URLs?
It depends on whether shoppers search for the variant. Google supports both path and parameter-based variant URLs, so separate variant URLs can be legitimate. For optional variant parameters, its guidance uses the parameterless product URL as the canonical. If a colour or size is simply a choice on one product page, one canonical product URL with consistent internal links is usually simpler to maintain.
Does a canonical tag fix duplicate collection pages?
Not in the sense people usually mean. A canonical states which of several similar URLs you prefer as the representative; it does not make two weak collection pages useful. If two collections serve the same shopper with substantially the same products, decide whether to merge or differentiate them. If they are genuinely different pages that only share a template, the work is editorial rather than canonicalization.
My supplier's description is on hundreds of other stores. Does that hurt me?
It is not a same-site duplicate-URL problem. It may be a differentiation problem if the page adds little beyond text available from many retailers. Check whether it adds store-specific sizing or compatibility detail, useful comparisons, original images, or customer evidence. StoreAuditLab can flag that condition in the reviewed sample but does not write product copy.
How much duplicate content is acceptable?
Google's current public guidance gives no acceptable percentage. A more useful test is per URL: is this address worth crawling, worth indexing, and worth a shopper's time? Ordinary variant, parameter, and collection-path duplication can be fine when signals consolidate consistently. Near-identical collections or templates that erase page-specific value need a page-level decision, not a percentage target.
When you want the sample reviewed for you
StoreAuditLab reviews up to 10 representative public URLs by hand, adds up to 10 Ahrefs-backed keyword and page opportunities, and delivers a prioritized plain-English PDF in 48 hours. $100 one time, fixed scope, no implementation upsell. Canonical, variant, and template duplication findings are reported with the evidence and the affected page type, not as a claim about what caused a traffic change.
Get the $100 audit →