Summary: There is no word count that ranks. Google has said word count is not a ranking factor, and the pages that rank are the ones that answer the search fully, whatever that takes. In practice that means a narrow question might need 500 words and a complete guide 3,000. Set the length by looking at what already ranks for your phrase, match or slightly exceed its usefulness, then cut rather than pad.
Ask ten people how long a blog post should be and you will get ten numbers, most of them from studies that are several years old. The honest answer is that length is a result, not a target: a post is the right length when the question is answered and no longer. This article explains why length seems to matter, what the studies actually found, what Google has said, and how to decide the length of any post in two minutes, with what we see across the sites we work on.
Word count is not a ranking factor. Google’s John Mueller has said so directly more than once. But longer posts do tend to rank better on average, for reasons that have nothing to do with the count itself.
A post that covers a question fully is usually longer than one that does not. Someone searching “how to start a vegetable garden” has a dozen sub-questions (soil, what to plant, when, watering, pests, common mistakes), and the page that answers them all is both longer and more useful. Length is a side effect of completeness.
A fuller post naturally uses the related terms and phrasings that help Google understand what it covers, and it can rank for several related searches rather than one. A short post on “best running shoes” mentions a few; a complete one covers road, trail, cushioning, pronation, budgets, and picks up searches for each.
A well-structured longer post keeps a reader on the page and moving through it, which Google reads as a sign the page answered the search. A long, badly structured post does the opposite.
Definitive pieces earn links and citations. Nobody links to a 400-word summary of something a dozen other sites have covered. That is an argument for depth, not for padding.
Two studies get quoted everywhere. Backlinko’s 2020 analysis of over 11 million search results found the average first-page result contained 1,447 words. HubSpot’s 2021 analysis of its own blog found posts of 2,100 to 2,400 words performed best in search. Both are real and both are routinely misread.
They are averages across every topic on the web. The average hides that a recipe ranks at 600 words and a software comparison at 4,000, and that the long pages ranked because they answered complex questions, not because they were long. Both studies are also several years old, from before AI Overviews began answering short questions directly in the results, which has made padding a short answer into a long page less useful than it ever was.
Use them for one thing only: as evidence that thin pages rarely rank for competitive questions. They tell you nothing about how long your post should be.
No. Google prefers the page that best answers the search. For a question with a short answer, a concise page beats a long one, and increasingly the answer is shown in an AI Overview before anyone clicks. For a question with a long answer, a short page cannot compete because it does not answer it. If the top results for your phrase are all 2,000 words or more, that tells you the question is big; a 600-word post will struggle. If they are all short, a 3,000-word post is answering questions nobody asked.
These are the ranges we typically see rank, by what the post is trying to do. They are observations, not rules.
| Type of post | Typical range | Why |
|---|---|---|
| Direct answer to a narrow question (“does VAT apply to X”) | 300 to 700 words | The answer is short; padding it hurts. Much of this traffic now goes to AI Overviews, so write the answer to be cited. |
| How-to or step-by-step | 800 to 1,500 | Enough to cover the steps, the tools and the mistakes without drifting |
| Comparison or “best of” | 1,200 to 2,500 | Each option needs a fair treatment and a recommendation by situation |
| Complete guide or pillar | 2,000 to 4,000 | Covers the subject and its sub-questions; links down to supporting posts |
| Cost or pricing guide | 1,500 to 2,500 | Ranges, the factors that move them, and examples |
| Original data or study | 1,000 to 3,000 | As long as the findings and method need; charts and tables do much of the work |
| News or update | 300 to 600 | Say what changed and what to do; nobody wants more |
Our guide to blog SEO covers the rest of the process, and our guide to on-page SEO covers the page elements.
When we review a site’s blog against Search Console, the posts that earn traffic are rarely the longest. They are the ones whose length matches their question: short direct answers that got cited, and long guides that owned a subject. The posts that earn nothing fall into two groups. The first is the 500-word post on a subject that needed 2,000, usually written to a schedule. The second is the 2,000-word post on a subject that needed 500, usually written to hit a word count. The fix for both is the same: decide what the question needs before writing, and merge or cut afterwards. Fewer posts, each the right length, outperform more posts at any length; our article on whether blogging helps SEO covers why. Our SEO packages include a content review that catches both.
AI Overviews and AI assistants have changed the calculation at the short end. A question with a two-sentence answer is now answered in the results, and a 1,500-word post written to rank for it gets fewer clicks than it used to even when it ranks. The response is not to write longer; it is to write the answer clearly enough to be the cited source, and to put the depth into the questions that need it. Length follows the question in both cases.
No. Google has no minimum. A page needs enough words to answer its question; for some questions that is a few hundred.
On average across the web, longer pages rank better, because they tend to answer bigger questions more fully. For any particular question, the best answer ranks, and that is often not the longest.
Only if they are short because they left the question half-answered. Adding words to a post that already answers its question makes it worse. Check what ranks for its phrase first.
The same as before: as long as the question takes. What has changed is that short-answer questions are increasingly answered by AI in the results, so the posts worth the most effort are the ones with questions that need depth.
See also, our complete guide to SEO for UK businesses.
Get practical SEO and digital marketing tips that help your business win more customers from Google - straight to your inbox, no jargon.
By subscribing you agree to receive marketing emails from Studio 36 Digital. Unsubscribe any time. See our Privacy Policy.