Thin content is any page that adds little or no unique value for the person who landed on it: too short, too shallow, duplicated, auto-spun, or stuffed with keywords instead of substance. Google has no word-count threshold. What it measures is whether the page satisfies intent better than the alternatives. When it doesn’t, the page either never ranks or quietly drags down the topical authority of everything around it.
Thin Content
Thin content is web page or site material that provides little to no unique value to users: short, shallow, duplicated, auto-generated, or keyword-stuffed copy that fails to satisfy search intent and underperforms in search.
Why Thin Content Hurts More Than It Used To
In the AI Overviews era, “good enough” pages are the first casualties. When Google can synthesize a direct answer from three authoritative sources, a 250-word page that restates the obvious has nothing to contribute and gets skipped, by the model and by the user. Thin content fails on two fronts at once: it doesn’t earn its own ranking, and at scale it signals a low-quality site, which is exactly what Google’s core ranking systems demote. (The separate helpful content system was folded into them in March 2024; the target stayed the same.)
Thin pages also dilute your site’s topical authority. A cluster of 40 solid pages plus 60 filler pages reads as a 60%-filler site. Pruning the filler often lifts the survivors, the most counterintuitive, most reliable win in a content audit.
Word count is a symptom. The diagnosis is whether the page answers the query. A 300-word answer that fully resolves the query is not thin. A 2,000-word page that buries the answer in fluff is.
The 6 Types of Thin Content (and How to Fix Each)
Most thin content falls into six recognizable patterns. Diagnose the type first; the fix is completely different depending on which one you’re looking at.
Short or shallow pages
Problem: Extremely brief pages that lack depth, context, examples, or actionable information (e.g., single-paragraph articles, minimal product descriptions).
Fix: Expand with useful details, data, visuals, user-intent-focused sections, FAQs, and real-world examples. Don’t pad to a number. Add the specifics a reader actually needs to act.
Duplicate content
Problem: Pages that repeat the same text across multiple URLs or copy content from other sites.
Fix: Consolidate or canonicalize duplicates, rewrite unique content, use 301 redirects or rel="canonical", and attribute or remove scraped material. Watch Search Console for Duplicate without user-selected canonical flags; they tell you exactly where Google is confused.
Low-quality automatically generated content
Problem: Machine- or template-produced pages with little human editing that read unnaturally or add no value (e.g., spun articles, bulk-generated category pages, and ungoverned AI output published at scale).
Fix: Rework with human oversight, add unique value, merge low-value pages, or remove them entirely. AI is fine as a drafting tool; AI as a publishing button is how you mass-produce thin content. Since March 2024, Google’s spam policies call this scaled content abuse: many pages made mainly to rank, with little value added, whether a model, a template or a room of freelancers wrote them.
Doorway and gateway pages
Problem: Pages created solely to rank for narrow keyword variations that funnel users to a single destination without unique content.
Fix: Merge doorway pages into comprehensive, user-focused pages; remove or redesign pages so each serves a distinct user need.
Thin affiliate/monetized pages
Problem: Pages that exist primarily to show affiliate links or ads with minimal original content or value.
Fix: Add original reviews, comparisons, firsthand experience, pricing data, multimedia, and clear buying guidance; disclose affiliations. This is where E-E-A-T (especially the first E, Experience) does the heavy lifting.
Poorly optimized paginated or parameterized content
Problem: Numerous near-identical pages created by pagination, filters, or URL parameters that cause thin variants (e.g., the same product list sorted differently).
Fix: Use canonical tags, noindex where appropriate, consolidate variations into single comprehensive pages, and govern faceted URLs with meta robots and x-robots rules. Search Console’s URL Parameters tool was retired in 2022, so those rules have to live on your site.
Thin Content Examples
What thin content looks like in a real crawl:
| Example | Type | Why it is thin |
|---|---|---|
| 300 city pages, identical except for the town name | Doorway / auto-generated | Swapping one word adds nothing a searcher in that town needs |
| A product page carrying only the manufacturer’s description | Duplicate | The same text sits on every retailer that sells the product |
| A “best X” post that lists affiliate links with one line each | Thin affiliate | No testing, no comparison, no reason to trust the picks |
| A 400-word post restating the top three results | Short / shallow | It offers nothing the pages it copies do not already say |
/shoes/?sort=price indexed next to /shoes/ | Parameterized | The same list in a different order |
| Tag and author archives holding one or two posts | Short / shallow | A list of links with no content of its own |
| Hundreds of unedited AI articles published in a week | Auto-generated | Scaled content abuse, in Google’s own policy terms |
How to Identify Thin Content
You can’t fix what you can’t see. Run these checks against a full content inventory before touching anything:
- Analytics: pages with very low average engagement time, high bounce rate, and near-zero conversions.
- Search Console: pages with impressions but few or no clicks, pages losing rankings over time, and pages piling up as “Crawled, currently not indexed” or “Discovered, currently not indexed” in the Page indexing report.
- Site crawl / duplicate checker: clusters of near-duplicate titles, meta descriptions, or body content.
- Content inventory: pages below a minimum depth threshold, or missing key elements (original assets, schema, CTAs).
- Manual sampling: review top pages for usefulness, originality, and alignment with intent; a human still catches what crawlers miss.
- Tools: Screaming Frog, Sitebulb, Semrush, Ahrefs, Google Search Console, and a plagiarism checker.
Fix, Consolidate, or Remove: The Decision Matrix
Every thin page gets one of three verdicts. Decide based on traffic, intent coverage, and business value. Leave sentiment out of it.
| Verdict | When to use it | Action |
|---|---|---|
| Improve | Page targets real demand and matches intent, but is shallow | Expand with original value, assets, and structure; re-promote |
| Consolidate | Multiple thin pages compete for the same intent | Merge into one comprehensive page; 301 the rest |
| Canonicalize | Near-duplicates from pagination, filters, or syndication | Set rel="canonical" to the primary; noindex only variants that should never appear in search, and never on a URL that also carries a canonical |
| Remove / noindex | No demand, no value, no path to either | Noindex or remove; 301 where relevant |
Prioritize by impact. The fastest win is almost always a high-impression, low-click page that already ranks on page two: it has demonstrated demand and just needs the depth to convert impressions into clicks.
Preventing Thin Content at Scale
Remediation is expensive; prevention is a process. The teams that never accumulate thin content bake quality into the workflow:
- Content standards: define required sections, depth, and user outcomes (word counts are secondary).
- Editorial briefs: every page maps intent, target keywords, competitors, required assets, and success metrics before a word is written.
- Templates with requirements: enforce value-add sections (unique intro, data, FAQ, CTA) so a pillar page can’t ship half-built.
- Governance: a recurring audit and pruning schedule, with named owners per topic.
- Limit auto-generation: never publish bulk-generated pages without unique, data-rich sections per page.
- URL and parameter rules: restrict indexing of faceted and parameterized URLs.
- Train for E-E-A-T: writers learn intent, original research, and firsthand experience, which matters most on YMYL topics.
- Test before scale: pilot a new page type and measure before mass-producing it.
That last point is the whole discipline behind programmatic SEO done right. The failure mode of pSEO is shipping 10,000 boilerplate pages with only a city name swapped: textbook thin content. Our Lab study of 12.3 million programmatic pages on nine large sites found 28.8% get exactly zero estimated organic traffic and 96.7% get 100 visits a month or fewer. We engineer the opposite: genuinely unique data in every template of a programmatic SEO build, so scale and quality move together.
When Thin Content Becomes a Business Problem
A blog with 40 posts and five weak ones is an afternoon of merging and rewriting. Use the matrix above and keep your budget.
Thin content costs money when it is a page type. Hundreds of location, tag or product pages with impressions and no clicks. A Page indexing report filling up with “Crawled, currently not indexed.” A sitewide drop after a core update, or a thin-content manual action in Search Console. At that point every URL needs a verdict from its own data, and the template that keeps producing thin pages needs fixing.
On a fractional SEO engagement, the two founders make those calls and fix the templates. On the Prawomat rebuild, 2,646 WordPress URLs became about 1,370: we cut 418 news posts and 252 court pages with no clicks and no backlinks, and rebuilt the 127 court pages that earned clicks or impressions.
Frequently Asked Questions
What counts as thin content in SEO?
Thin content is any page that adds little or no unique value for the searcher: extremely short pages, duplicate or scraped copy, auto-generated or doorway pages, thin affiliate pages, and near-duplicate parameterized URLs. The test is whether the page satisfies intent better than the alternatives. Word count is not part of that test.
What are examples of thin content?
City or location pages that differ only by the place name, product pages that reuse the manufacturer’s description, affiliate roundups with one line per product, short posts that restate what already ranks, filtered or sorted URLs indexed as separate pages, near-empty tag archives, and unedited AI articles published in bulk.
Does thin content cause a Google penalty?
It can. Google has a manual action called “Thin content with little or no added value,” reported in Search Console’s Manual Actions report, which covers auto-generated pages, thin affiliate pages, scraped content and doorway pages. Most thin-content losses are algorithmic, though: Google’s core ranking systems demote low-value pages, and at scale that can drag down a whole site. Since March 2024, mass-produced low-value pages also fall under the scaled content abuse spam policy.
How many words does a page need to avoid being thin?
There’s no minimum word count. A focused 300-word answer that fully resolves a query is not thin. A padded 2,000-word page that buries the answer is. Google measures whether the page satisfies intent and adds unique value: depth, originality, and usefulness. Length is not on that list.
Should I delete thin content or improve it?
Improve pages that target real demand and match intent; consolidate multiple thin pages competing for the same intent and 301 the duplicates; canonicalize near-duplicates from pagination or filters; and noindex or remove pages with no demand and no path to value. Prioritize by traffic, conversions, and strategic importance.
Is AI-generated content automatically thin content?
No. Google judges content by its quality and usefulness, whatever produced it. AI output becomes thin when it’s published unedited at scale with no unique data, experience, or oversight. Used as a drafting tool with human review and original value added, AI-assisted content can rank perfectly well.