Content Length for SEO: What Counts and How to Fix It

What is content length in SEO?

Content length in SEO means the number of words in the main content of a page: body copy, headings, list items, table text and captions. Navigation menus, footer links, cookie banners and sidebar blocks repeated on every URL are boilerplate, so leave them out when you compare pages.

Not the HTTP header: search for the bare phrase and you will mostly meet something else, the Content-Length header. It states the size of a response body in bytes and is defined in RFC 9110, so it has nothing to do with editorial depth. The size of the HTML document in kilobytes is another measurement again, which our page size check covers.

Length is also not content depth. Depth describes how fully a topic is handled, while length is only the visible text that coverage produces, so a padded page can be long and still shallow. Text baked into an image does not count as copy.

Word count, character count and page weight

• Word count. The words a reader sees, and the editorial measure for planning and comparing articles.
• Character count. Letters and symbols. It governs title and description limits and is how raw HTML length is counted.
• Page weight. File size in bytes, or everything the page loads. It matters for speed, not for how thorough your writing is.

Does content length affect SEO?

Indirectly, yes. Directly, no, so the ideal content length for SEO is not a fixed number. Google's own advice on creating content includes a self check that asks whether you are padding to hit a word count you heard Google likes, and it says no such preference exists. John Mueller has said the same in public, and the position is stated in Google's helpful content guidance.

So why do long pages so often rank well? A 2,000 word piece of long form content has room to answer follow up questions that show up in People Also Ask, so it matches more long tail keywords and earns more links. The coverage is the more likely reason they rank, not the length itself.

We measured this for the very query behind this guide. The ten ranking guides ran from roughly 670 to 3,400 words, with a median near 1,900. No single length won, because intent sets the bar.

The real damage happens at the thin end. Shallow content with no original content or information gain can be treated as low value, and Google lists thin pages that add little or nothing new as a manual action type. That action targets thin affiliate pages, scraped or low quality guest content and doorway pages, so a short page that fully answers its query is not the target.

Padding has its own cost, since readers leave when the answer sits under pages of preamble. A 400 word definition can easily beat a 2,500 word essay for a query like what is a canonical tag, and a featured snippet or AI Overviews answer tends to lift one self contained passage, whatever the size of the page around it. For the wider picture, see our on-page SEO basics.

How SEO Site Checker tests content length

Our report opens your URL in headless Chromium, which runs the page's JavaScript like a visitor's browser. The Content Length Test then reads the rendered HTML document, so scripts have already done their work.

Exact rule: the check counts the characters in the rendered HTML, markup included. Under 800 characters warns, and 800 or more passes.

• Pass: the rendered document is 800 characters or longer.
• Warning: the rendered document is under 800 characters, a near empty page.
• Not tested: visible word count, text quality, topical coverage and how you compare with ranking pages.

Here is the part to know before you trust the number. The sentence in the report talks about words on the page, but the figure behind it is a character count of the HTML, and tags, inline scripts and inline CSS all count toward it. Almost any real page clears 800 characters, and we'd rather say so than imply more than the tick measures.

A warning usually means a blank page, a bare maintenance notice, a plain text or JSON response, a redirect to an almost empty URL, or a script driven app that failed to render. Larger interstitials such as bot challenges, error pages and login walls often run past 800 characters of markup, so they tend to pass without holding useful copy. The AI Search Readiness panel reuses this number for its content depth line. It applies its own cutoffs to that same character count, so treat that line as a floor too, not a measure of depth.

What the three results mean

A pass needs 800 characters or more and shows a green congratulations line with the count, with no fix button. A warning, under 800 characters, shows a red alert with the count plus the How to Fix button. A Failed state at zero characters shows the same red alert and button, but you will almost never see it, because an empty fetch stops the audit with an error first.

Why a pass is only the floor

A pass screens out empty pages and nothing more. A truly bare page, such as an empty stub or a one line maintenance notice, warns, which is correct. But a coming soon template with inline CSS, or a blank app shell stuffed with inline scripts, can pass with no real copy at all. Treat green as permission to move on, then count your words and compare them with the top ranking pages.

Run the free SEO score report

How many words does each type of page need?

There is no official table, so treat the ranges below as working guidance rather than Google rules. The first four rows come from the text length checks in Yoast SEO, a popular WordPress plugin, and the last three are commonly cited ranges that vary by source.

The spread we measured earlier makes the same point. Definitions and conversions can win with a few hundred words, while comparisons and tutorials need far more, because the searcher wants steps, examples and edge cases.

Worked example: suppose the top three results for your keyword run 700, 1,400 and 2,100 words. Don't copy the middle one. List the sub questions each result answers and plan around the combined list, then write as many words as that list needs.

How to check the content length of any page

The automated line gives you a floor. For real word counts, use a short routine that works on your pages and your competitors' alike.

1. Run the report. Enter the URL in the free report and read the Content Length Test line to rule out a near empty page.
2. Count visible words. Open Chrome DevTools, switch to the Console tab, and paste the snippet below, or use any word counter on the copied text. If Chrome blocks the paste, type allow pasting and press Enter first. For unpublished drafts, use your editor's counter instead, such as the Google Docs word count under Tools or the information panel in the WordPress block editor.
3. Benchmark the results. Run the same snippet on the top three to five ranking pages and note which sub questions every page answers. Your own average word count for that one query is the only average that matters.
4. Write down the gap. Record your number plus what the others cover and you do not, which becomes your rewrite outline.

Console snippet to count words

document.querySelector('main').innerText.trim().split(/\s+/).length

This counts the main content only, so navigation and footer text stay out of the number. If the console throws Cannot read properties of null, the page has no main element, so swap main for article or body. Because innerText skips text hidden in collapsed tabs or accordions, open them first, or swap in textContent, which also counts hidden script and style text.

Curious how the report's own figure compares? Run document.documentElement.outerHTML.length in the same console. It prints the characters in the rendered HTML, markup included, which is close to what our check counts but is not a word count.

How to fix thin content and a low word count

Work from the cause to the cure, because adding words to a page that fails to render wastes time.

1. Rule out a rendering problem. Load the page in a normal browser first. In Search Console, URL Inspection and then View crawled page shows the HTML Googlebot received. If the copy is missing there, or only the crawler sees a blank page, look for a failing script, a bot challenge or an error response, and fix delivery before you write anything.
2. Decide the page's job. Name the one search intent it serves, and confirm it deserves its own URL instead of a section on another page.
3. List the gaps. A content gap analysis is simple here: note the questions the top results answer that you do not, plus the related questions people ask, skipping any that belong on a different page. Turn each into a heading with a direct answer, and use our keyword usage check to see which terms the page already covers. Keep the hierarchy tidy, as our heading structure guide explains.
4. Add what only you have. Include first hand examples, original data, screenshots with alt text and steps you actually tested.
5. Merge overlapping pages. Combine thin pages on one topic into one stronger URL, redirect the old addresses with a 301 redirect and update internal links.
6. Prune the hopeless ones. Start with a quick content audit: export your URLs and mark the empty archives, test pages and duplicates, which should get a noindex tag or be removed. Our noindex tag guide shows how.
7. Re-test and request a recrawl. Run the report again, then use the URL Inspection tool in Google Search Console to ask for a recrawl.

If you suspect a manual flag, check the Manual Actions report in Search Console and read Google's thin content manual action help text, which describes what counts as thin and how to request a review.

Platform notes for WordPress, Shopify and Wix

• WordPress. Open Pages, choose Edit, and write the copy in the WordPress block editor. If your homepage is empty, check under Settings, then Reading, that the static page you picked contains text.
• Shopify. Under Products, open an item and fill the description box. Do the same under Products, then Collections, and write standalone pages such as shipping or about under Online Store, then Pages.
• Wix and Squarespace. Add text elements or text blocks for your copy, and avoid pasting a screenshot of it, because a crawler cannot read text inside an image.

Content length mistakes to avoid

Most length problems start when someone chases a number, which leads to content padding and keyword stuffing instead of serving the reader.

• Copying a competitor's length. Matching their 2,100 words while skipping the comparison table they lead with gives you the same size and less value. Match their coverage instead, then add what they missed.
• Using one target for every page. A glossary entry stretched to 1,500 words buries a two sentence answer.
• Burying the answer. A 600 word preamble before the point makes readers hunt for what they came for.
• Mixing two intents. A page that tries to be both a how to guide and a product pitch ranks for neither, so give each search its own URL.
• Counting boilerplate. A page showing 900 words, 400 of them repeated menus and footer links, is really a 500 word page.
• Trusting a script to deliver the copy. If the main text only appears after a script fetches it from an API, a crawler may see an empty page. Server side rendering or prerendering the main copy fixes that, since the copy then arrives in the first HTML response.

Content length best practices

• Start from intent and the results. Review the top results before you outline, look for gaps, then write until the question is fully answered and stop, even if the count looks low.
• Lead with the answer. Put the direct answer near the top, then expand in clearly headed sections for readers who want the detail.
• Update instead of duplicating. Refresh an old page rather than publishing a near copy. Our SEO for bloggers page shows how this fits a content routine.
• Add wordCount only if you want it. You can add an Article block with a wordCount value in JSON-LD, using the Schema.org wordCount property. Google's Article documentation does not list it, so treat it as plain documentation for other consumers and not a ranking lever.
• Re-check key pages. Run the report again after a redesign, since a broken script can empty a page that was fine last month.