Finding and fixing thin and duplicate content on your website

Fibvo editorial · 12 February 2026 · 4 min read

A transport company has forty pages targeting truck hire in different cities. Each page says the same thing with the city name swapped. An industrial supplies store sells two hundred products, yet Search Console lists thousands of URLs, most ending in ?sort= or ?colour=. The two sites look different but share one problem. Search engines are spending time on pages that add nothing, and the pages that matter are harder to pick out.

Thin and duplicate are different problems

Thin content is a page with little value of its own. Duplicate content is the same, or nearly the same, content at more than one URL. They overlap, but the fixes differ.

• Thin: city pages with only the place name changed, tag archives holding one post, product pages with nothing but a manufacturer's description and short posts that answer no real question.
• Duplicate: http and https versions, www and non-www, trailing-slash variations, the same product reachable through two category paths and URLs created by filters, sorting and tracking codes.

Ordinary duplication on your own site is usually filtered, not penalised. Google groups the copies and chooses one to show. The cost is that it may choose the wrong one, and links and other signals end up split across several addresses. Deliberate copying is different. Google's spam policies do act against content scraped or copied from other sites and published to manipulate rankings, and near-identical city pages made only to catch searches can be treated as doorway pages.

Google usually filters ordinary duplicate pages rather than penalising them. The cost is that it stops listening to you about which one matters.

Where to look for them

Start with the data Google already gives you, then fill the gaps with a crawl.

• In Search Console's Page indexing report, open 'Duplicate without user-selected canonical', 'Duplicate, Google chose different canonical than user' and 'Crawled - currently not indexed'. Export each list.
• Crawl the site and sort pages by word count. Many crawlers also flag near-duplicate pages.
• Filter the crawl for URLs containing a question mark to see how many parameter versions your navigation creates.
• Compare your crawl with the Pages tab of the Performance report over the longest date range. Indexable pages missing from that tab have earned no impressions at all.
• Search Google for a distinctive sentence from a page in quotation marks to find other sites carrying the same text.

Improve, merge, canonicalise or noindex

For each problem URL, ask two questions. Does the page serve a searcher that no other page serves? Does it need to exist for visitors even if it should not appear in search? The answers lead to one of four fixes.

• Improve it when the page has a real purpose. A city page earns its place with the local depot address, routes served from there, vehicle types based there, typical transit times to nearby industrial areas and a named local contact.
• Merge it when several weak pages cover one topic. Combine the best material into a single page, 301 redirect the others to it and update internal links to the new URL.
• Canonicalise it when duplicates must exist for visitors, such as a product listed in two categories. Add a rel=canonical tag on each version that points at the preferred URL.
• Noindex it when a page helps visitors but not searchers, such as internal search results or thin tag archives. Use a robots meta tag and leave the page crawlable.

Remove a page with a 404 or 410 status only when it has no traffic, no links and no purpose.

Rules for canonical and noindex tags

Both tags are simple to add and easy to misuse. These rules prevent most mistakes.

• A canonical tag is a strong hint, not a command. Google follows it most reliably when internal links, the sitemap and redirects all point at the same preferred URL.
• Give every indexable page a self-referencing canonical using its full https address.
• Do not point paginated pages at page one. Page two of a category shows different products and should name itself as canonical.
• Do not block a noindexed page in robots.txt. Google has to crawl the page to see the tag.
• Do not put noindex and a canonical pointing elsewhere on the same page. The two signals contradict each other.
• Google does not support noindex rules inside robots.txt. Use the meta tag or an X-Robots-Tag HTTP header.

Bring parameter URLs under control

A few sort orders, colours and sizes can multiply one category into dozens of URLs. Google has retired the URL Parameters tool from Search Console, so control now sits on your own site.

• Tracking parameters such as utm_source: keep them off internal links entirely, and make sure the canonical tag shows the clean URL even when a parameter is present.
• Sort orders and view options: canonicalise to the unsorted category URL.
• Filters buyers genuinely search for, such as stainless steel ball valves: give them a clean static URL, a unique heading and some written content, and allow indexing.
• Filter combinations nobody searches for: keep them out of crawlable links where possible. If they still cause heavy crawling, block the pattern in robots.txt, accepting that Google cannot read canonical tags on blocked URLs.

Make changes in batches and watch the Page indexing report over the following weeks. Duplicate counts should fall as Google recrawls. If they do not, check whether a plugin, the sitemap or old internal links are still generating the URLs you meant to retire.

---

Written by Fibvo — Bengaluru and Erode, India — working with clients in other time zones. Questions, or think we have it wrong? Tell us.