How to fix thin content

Fix thin content by doing one of three things to each affected page: merge near-duplicates into a single strong page and 301 the rest, add the specifics a searcher actually needs, or noindex pages that exist to serve navigation rather than search. Thin means a page that fails to satisfy the visit — not simply a low word count — so the fix is always about what is missing rather than about hitting a length target.

Last updated 2026-07-28

What counts as thin content?

A page that adds little or no value beyond what already exists elsewhere. Google's own wording is "little or no added value", and the canonical examples it lists are scraped content, auto-generated text, doorway pages, thin affiliate pages and near-duplicates of your own other pages.

Note what is absent from that list: shortness. A 150-word page that answers one question completely and correctly is not thin. A 2,000-word page assembled from restated introductions and competitor paraphrase is, however long it runs.

The practical test is whether the visit ends. If someone arrives from a search, reads the page, and goes back to the results to try the next link, the page was thin for that query regardless of what the word counter says.

Is there a minimum word count for ranking?

No, and Google has said so repeatedly and unambiguously. There is no threshold, no target, and no bonus for length — the correct length is whatever fully answers the query, which for some intents is 200 words and for others is 2,000.

Word count is still worth measuring in an audit, but as a flag rather than a rule. A cluster of pages under about 200 words reliably indicates something structural: placeholder pages that were never finished, empty category listings, paginated archive stubs, or templated variants generated from a spreadsheet. The number is not the problem; it is the tell.

Treat it the way you would treat a Lighthouse score — a diagnostic that points at where to look, not a target to optimise directly. Padding a 180-word page up to 800 words changes the metric and makes the page worse.

Which thin pages should be fixed, merged, or noindexed?

Sort them by cause first, because each cause has exactly one right answer. Near-duplicates get consolidated, genuinely useful but shallow pages get expanded, utility pages get noindexed, and mass-generated pages with no unique value get removed.

How do you add value rather than words?

Add the things a reader would otherwise have to go and find somewhere else: the exact number, the actual markup, the real threshold, the specific failure mode. Every sentence that saves the reader a search is value; every sentence that restates the heading is padding.

• Answer the follow-up question a reader would ask next, in the order they would ask it. Most thin pages stop at the first answer.
• Include something only you can provide — original data, your own screenshots, a real example from work you have done.
• Give a concrete procedure: steps, code, a command, a table that saves the reader assembling it themselves.
• State the numbers. "Keep it short" is padding; "120 to 160 characters, truncated at about 920 pixels" is content.
• Say what does not work and why. Failure modes are among the hardest things to find elsewhere and the most useful when found.
• Cut the introduction. Pages that open by explaining that the topic is important lose readers before the first fact.
• Delete padding aggressively at the end. Length added for its own sake makes a page worse on every measure that matters.

Does thin content trigger a penalty?

Rarely a manual penalty, but routinely a ranking problem. Manual actions for thin content do exist and appear in Search Console when applied, yet they are uncommon compared with the far more ordinary outcome: the page simply never ranks, because nothing about it improves on the results already there.

The exception is scale. Google's spam policies specifically name scaled content abuse — generating large numbers of pages primarily to manipulate rankings rather than to help people — and enforcement here is algorithmic and site-wide rather than page-by-page. A few hundred templated location pages built by find-and-replace is the classic trigger.

The knock-on effect matters more than the penalty question for most sites. A large body of thin URLs consumes crawl budget, dilutes internal link equity across pages that will never rank, and drags the average quality of the site down for the pages you actually care about.

How do I find the thin pages on my site?

Crawl the site, sort every URL by visible word count ascending, and read the bottom of that list as groups rather than individual pages. Thin pages cluster by template almost without exception, so twenty rows usually represent one decision.

Audra measures visible word count per crawled page and flags unusually low values, reporting them next to the title, meta description and robots columns for the same URL. That combination is what makes the diagnosis quick: a block of pages sharing a template, a title pattern and a low word count is a consolidation candidate, and one sharing a low word count with an existing noindex is already handled. Audra also runs its own answer-engine optimisation checks, which look at whether a page states its answers plainly enough to be extracted — the failure mode thin pages share with padded ones.