Word count is not a word count seo ranking factor, at least not by anything Google has said. Google says the length of a page alone does not matter for ranking and that it has no preferred word count, so the claim that a post needs a certain number of words to rank is debunked as a direct rule. If you have ever been asked "does word count matter for SEO?" in a planning meeting, that is the short answer, and this piece walks through the evidence behind it. The one thing the evidence does not say is that what a page contains is unimportant. A longer answer can serve a reader better, and that is a different claim from a number that Google rewards.
We're grading the evidence as debunked for the specific idea that Google applies a preferred word count or a minimum length. The strongest sources are Google's own guidance, and the weaker one is an industry study that gets quoted a lot. We'll go through both, dated and attributed, so you can tell which is which.
What the word count claim says and where it came from
This content length myth comes in a few versions. One says Google directly rewards a page for having more words. Another says a post has to clear some threshold, often 1,000 or 2,000 words, before it can rank. The third is a softer version that shows up in content briefs, where a tool or a manager sets a target and the writer is told to hit it.
It helps to pull three different ideas apart, because most of the confusion comes from mixing them.
- A direct word count rule. Going from one length to another earns an advantage because of the count itself. Google's published guidance rejects this.
- Useful coverage. More words can carry an explanation, original information, or an answer that a shorter version left out. The usefulness may matter to a reader, and the number of words is not a substitute for checking it.
- A measured association. The average length of pages that already rank, or a link between long articles and backlinks, describes a sample. It can't show that Google rewarded the length.
So the defensible version of the claim is narrower than "length never matters in any way." Length alone is not a Google ranking criterion, and Google has not given anyone a length to aim for.
The evidence that seems to support it
The number most people have seen comes from an analysis by Backlinko using Ahrefs data. Its methods document says the Google results were collected in the first week of January 2020. After cleaning, the study covered 1,183,628 keywords and 11,835,086 ranking URLs, and it looked at the top ten results for each keyword. The researchers scraped more than one million pages for on-page measures, and word count was one of them.
The figure that gets repeated is 1,447 words, the average length of a top-ten result in that sample. It's easy to see why it spreads. It looks like a target, and it's specific enough to put in a brief.
There are a few things to keep in mind before using it that way.
- It is vendor research. Backlinko and Ahrefs ran it, and it is not a disclosure from Google about how ranking works.
- It is old data with a newer date on the page. The search results were collected in January 2020. The Backlinko article was last updated on April 14, 2025, but that update date doesn't make the underlying snapshot current.
- It only looks at pages that already rank. A top-ten-only design describes what is in the top ten. It can't tell you what would happen if you padded one existing page.
- The study itself warns about causation. Its methods document says correlation does not necessarily imply causation and that the study can't assign causation to a factor.
Backlinko also says something that rarely gets repeated next to the 1,447 figure. It reports finding no direct relationship between word count and rankings. So the same study that gave people the average also says the average is not evidence of a rule.
The study notes that other industry research has linked longer content to more backlinks. That is an observation about links, and it doesn't show that adding length causes links or ranking gains for your page. If you hear "long content ranks because it earns links" stated as settled, it isn't, at least not from these sources.
What Google's own documentation says
Google Search Central is the primary source here, and there are two pages that speak to it directly. Both show a last-updated date of December 10, 2025.
The SEO Starter Guide says: "The length of the content alone doesn't matter for ranking purposes (there's no magical word count target, minimum or maximum, though you probably want to have at least one word)." The word to hold onto is "alone." Google is saying the count by itself is not the thing being measured. It is not saying that what the words communicate is unimportant. The bit about having at least one word is a light aside and not a threshold anyone can act on.
The guidance on creating helpful, reliable, people-first content is more direct. It lists questions to ask about content that is made mainly to attract search visits, and one of them is: "Are you writing to a particular word count because you've heard or read that Google has a preferred word count?" The answer printed next to it is "(No, we don't.)" That speaks to the exact habit of writing toward a number because you assume Google wants it.
The same page describes Google's automated ranking systems as built to prioritize helpful, reliable information made to benefit people, and not content made to manipulate rankings. It gives a set of self-assessment questions, and they are about substance. They ask whether the content offers original information, reporting, research, or analysis, whether it substantially describes the topic, whether it shows firsthand expertise where that's relevant, and whether a reader would come away having learned enough to reach their goal and had a satisfying experience.
Those are editorial questions, and they are where the real google word count quality conversation belongs. They are not a formula, and they aren't a promise that answering them well guarantees a position. The Starter Guide also tells creators not to just rehash what others have already written. It's worth reading those questions as a set of things to review on a draft, and not as a request to cover every nearby topic.
Why a short page is not automatically thin content
A lot of the worry about word count shows up in a low word count seo audit. A tool crawls the site, flags every page under some number of words, and labels the list "thin content." It's natural to read that as a warning from Google, and it isn't one.
Google's Search Console Help does list "Thin content with little or no added value" as a possible manual action. It does not attach a word count threshold to that action. Google's spam policies talk about scaled content abuse in terms of producing many pages mainly to manipulate rankings and not to help people, usually with unoriginal content that offers little or no value. The discussion of thin affiliate pages is about copied product descriptions or reviews that add nothing of their own. Neither policy says that short posts are spam.
So a low count in an audit is a prompt to look at the page, and it doesn't tell you anything is wrong yet. Google's December 2025 guidance gives no numerical cutoff. A short page can fully answer a narrow question. A very long post can still repeat familiar material, skip the answer, or leave a reader needing to search again. Those are examples of how it can go, and not measured outcomes, so please don't quote them as data.
If a page really does carry a manual action, that's a separate situation. Google's advice there is to open the details in the manual actions report, fix the affected pages, and request a review. An audit label like "thin" doesn't mean you have one, and you can check by looking at the report and not by counting words.
If your team runs these audits regularly, it can help to pair the flag with a review of which pages are decaying or worth refreshing, since that tells you where an edit is likely to be worth the time. Our guide on running a content audit to find decaying pages covers that side of it, and it also helps to know how to flag pages for refresh in a repeatable way.
How to review a flagged page instead
When a page gets flagged, the useful question is what the reader was trying to learn and whether this page gives them the answer along with enough supporting detail. That is the google word count quality question restated in plain terms: the quality is in whether the answer is there, and not in how many words surround it.
You can work through it in a few steps.
- Write down the question the page is meant to answer. Use the words a reader would type or ask, and keep it to one sentence. If you can't, the page may be trying to do too many things.
- Read it as that reader would. Look for the answer near the top, then check whether the context you'd need to trust it or act on it is there.
- Ask whether the reader would have to search again. If they would, find the specific thing that's missing and add it.
- Look for material that only repeats or delays the answer. Cut it, or tighten it. Shorter is fine when the answer is already clear.
- Check the original information. Google's questions ask about original analysis and firsthand expertise, so if the page only restates what other pages say, that's the gap to fix, and it isn't a length gap.
A narrow factual question and a contested, many-sided claim don't need the same depth, and you can't judge either from its count. That's reasoning about what readers need, and it isn't evidence that either kind of page will rank, and it isn't a hidden length recommendation either. If you're looking for how long different kinds of posts tend to run, that's a separate question from this one, and this piece deliberately doesn't set a target.
It also fits with how search intent shapes SEO content, since the depth a page needs comes from what the searcher is after.
What to do differently on your team
If you lead a content team, the practical change is small, and it starts with the brief. Swap "how many words do we need?" for "what does this page have to explain for the reader's question to be answered?" Then judge drafts on how accurate and original the answer is, whether it supplies the context a reader needs, and whether they would have to go somewhere else to finish the task.
Add material when it answers a real question that's still open. Cut material that only repeats itself or slows the answer down. This is an editorial reading of the questions Google has published, and it's not a newly discovered ranking algorithm, so it's fine to say that plainly when someone asks where the rule came from.
A few habits make this easier to keep up.
- Take word count fields out of briefs as a requirement. You can leave a rough range as a planning aid for how much time a piece needs, as long as everyone knows it isn't a ranking input.
- Review flagged pages against the reader's question, not the tool's threshold. The refresh checklist is a reasonable way to keep those reviews short.
- Watch the quality-versus-volume tradeoff. Padding to hit a number takes time away from the pages that could use real work, and a simple content planning framework can help you decide where that time goes.
- Keep the E-E-A-T questions in the review. Firsthand expertise and original information are things Google asks about, and how E-E-A-T signals vary by intent gives you a way to think about them by page type.
If your pages also need to show up in AI answers, length works the same way there, in that it's the clarity and completeness of the answer that counts. We looked at that in how long an article should be to get cited, and a page that ranks well can still miss those citations, which we cover in why high-ranking pages still miss AI citations.
What would change this verdict
The verdict rests on what Google has said publicly, and that has limits. Google's statements establish that it has no preferred word count, but they don't disclose every internal consideration for every query. So it would be overreach to claim that extra explanatory detail could never help a page. Whether one particular longer revision serves one particular query better is a case-by-case question that these sources can't settle in advance.
Here is what would move the grade. If Google published guidance that named a minimum or target length, the debunked rating would need to change. A controlled experiment that isolated word count from everything else on a page and showed a consistent effect would count as well, and none of the sources here is that. The Backlinko study is observational, from January 2020, and by its own account can't assign causation. Until something stronger shows up, the safest reading is the one Google gives in its own words.
The takeaway for a marketing lead is that you can stop treating a word count seo ranking factor as a quality score, since the evidence doesn't support one. Look at whether each page answers the question it was written for, add what's missing, and leave the rest alone.
If you want a system that researches, writes, and links each article from your brand context, so that your review time goes to whether the piece answers its reader, you can start a free trial of DeepSmith.
