Thin content
Thin content is a page with little or no value of its own. For a small site, it wastes crawl, dilutes relevance and can pull the whole domain down. Here is how to find it, read it in Search Console, and choose to expand, merge or remove.
By Théophile Louvart, founder of Porteur · Updated 13 September 2026 · Markdown
What thin content means for a small site
A thin page adds little beyond what already exists on your site or the web. It might be a short stub, a template with filler, or a duplicate without a clear purpose.
On a small site, a few weak pages can hurt many strong ones. They split signals, confuse internal links and invite intent mismatch. Google’s helpful content signals are, as of 2026, part of core ranking. A pattern of unhelpful pages can depress the whole site.
Example: you publish ten near-identical pages for “/features/crm”, one per micro use case, each with 120 words and no data. None ranks. Your main “/features” also slips because your site looks shallow on the topic.
How to spot and measure it
- Search Console, Performance: sort by impressions then clicks. Pages with many impressions and near zero clicks are suspects. Open the queries and the page.
- Search Console, Pages report: look for “Crawled, not indexed” patterns on near-duplicate templates.
- Read the page itself: does it answer one clear search? Does it contain data, examples, screenshots, code, pricing, or a result you produced? If not, it is thin.
- Check cannibalisation: two pages both ranking around position 15 for the same query. Decide which should win. See also “Keyword cannibalization: two pages, one search, and how to fix it”.
- Check duplicates: if two URLs serve the same thing, pick a canonical. Cross check titles and headings for copy-paste patterns.
Ignore word count as a rule. A 200 word “/status” page can be fine. A 1,500 word fluff piece can be thin. Judge by original value to the reader.
Fix it: expand, merge or remove
Pick the primary page
Choose the URL that should own the topic. Example: keep “/guides/getting-started”, not five near copies under “/blog/quick-start-…”. Link to it from relevant pages.
Merge overlapping pages
Move the best parts into the primary page. 301 redirect the merged URLs to it. Update internal links to point to the winner.
Expand where you have something real to add
Add your data, tests, screenshots, code, FAQs and outcomes. Example: replace generic claims with a table of your plan limits from “/pricing”, or a short video.
Use canonical or noindex for unavoidable duplicates
If two URLs must exist, set the canonical to the preferred one. Use meta robots noindex for tag pages, feeds, or search results that add no value.
Remove dead ends
Delete orphan stubs that serve no user. 404 or 410 them. Replace doorway pages with a single intent-matched landing page.
A fixed page looks like this: “/integrations/slack” gains setup steps, screenshots, limits, a changelog, and links from “/docs”. Queries and clicks rise, and rivals stop outranking it for your brand plus “Slack integration”.
Spam policies and common traps
- Scaled content runs: thousands of city or keyword pages with the same template and thin swaps. If each page lacks unique value, expect it to lose.
- Third-party filler: hosting generic reviews or coupons on your domain to ride your authority. This risks site reputation abuse.
- Expired domains: buying an old domain and filling it with off-topic or low value content. Short term lift, long term risk.
- Programmatic pages: can work when each page holds its own data, eg live inventory by city. It fails when pages are near-empty shells.
- AI drafts: useful as a starting point. Publish only when you add your experience, data and examples. Do not scale blank pages.
Questions
It is a page with little or no value of its own. Think stubs, duplicates, doorway pages or filler that does not answer a clear search. Judge by usefulness, not length.
Google uses “thin content with little or no added value” to describe pages that add nothing new or helpful. As of 2026 this can limit how your whole site ranks, because helpful content signals sit in core systems.
No. Google does not penalise AI-assisted writing as such. It targets scaled pages made mainly to manipulate rankings and that lack value, no matter who or what wrote them.
There is no safe length. A short page can be strong if it answers the query with specifics. A long page is thin if it adds nothing new or useful.
Choose a preferred URL and set a canonical to it. If a page has no search value, use meta robots noindex. If two pages overlap, merge and 301 to the winner.
You may see short dips on removed URLs. Over a few weeks, stronger pages tend to gain. You also cut crawl waste and reduce cannibalisation, which helps the domain.
Sources
Check my site, free
Want a quick read on weak sections and overlap on your site? Run the free check from your URL, in about thirty seconds, and see three findings whole.
- Free check, no card
- Read-only, your own accounts
- Readable by your agent
Read next
- GlossaryScaled content abuse
- GuideProgrammatic SEO for a small product: page families that earn
- GuideSearch intent: what the results page tells you to build
- GuideInternal linking for a small site: which pages link to which
- GuideOrphan pages: finding the pages nothing links to
- GlossaryContent decay
- GlossaryDuplicate content
- GuideHow to fix duplicate content, case by case