Short posts get fewer backlinks, but no study shows a word count where links multiply. What Semrush, Backlinko and Ahrefs found, and which formats earn links.
Content length is linked to backlinks, but not in the way some roundups claim. The clearest finding is at the short end: in Semrush's study of more than 1,200,000 articles, posts of 300 to 900 words got 75% fewer backlinks than articles of average length. Longer posts tend to earn more links, but the big studies disagree on where that stops, and none shows a word count past which links multiply. Format matters too: some types of post earn more links than others. Here is what the data says, where a popular stat goes wrong, and how to size a post you want links for.
What the studies found
Semrush's study of top performing articles, is dated February 2, 2021. It "analyzed the visits, backlinks, and social shares of 1,200,000+ articles". Its key takeaway on length reads: "7,000+ word articles drive almost 4 times more traffic and 43% more shares than articles of average length (900-1,200 words). Meanwhile, short posts (300-900 words) gain 21% less traffic and 75% less backlinks than articles of average length (900-1,200 words)."
So in this study, the backlink figure is about short posts losing links. The "almost 4 times" figure is about traffic for very long posts.
Backlinko's study of 912 million blog posts, was run with BuzzSumo data and last updated on Feb. 19, 2019. It looked at length and links directly: "Content longer than 3000 words gets an average of 77.2% more referring domain links than content shorter than 1000 words." The same study found "94% of all blog posts have zero external links."
Ahrefs found a different shape. In its analysis of content length, based on a sample of around 900 million pages in its Content Explorer tool, Joshua Hardwick reports a strong positive correlation between word count and backlinks up to 1,000 words. Past that point the correlation turns strongly negative, and the average number of referring domains starts to decrease. Ahrefs' theory is that very long posts bury their link-worthy points, so fewer readers reach them and fewer link to them. Its advice is to focus "on including all link-worthy points as succinctly as possible."
Both are correlation studies. Semrush itself says long reads "likely perform better because they provide users with in-depth information on the topic" and not "just because they contain more words."
Which content formats earn the most links
Format shows up in the link data too. Backlinko's study found that why posts, what posts and infographics "receive an average of 25.8% more referring domain links than how-to posts and videos". List posts were the opposite. They were the top format for social shares, but "dead last in terms of getting backlinks from other websites". The study also found almost no correlation between shares and links, with a Pearson correlation coefficient of 0.078. A post that does well on social media will not necessarily earn links.
Backlinko's explanation for infographics is that they carry citable data and can be embedded in other people's posts. The study was last updated in February 2019. Treat the ranking as a direction, not a rule.
The stat that does not match the source
A claim that shows up in some stat roundups says Semrush found that posts over 3,000 words earn three or four times more backlinks. That is not what the 2021 study's text says. Its multiplier belongs to traffic at more than 7,000 words, and its backlink number belongs to short posts. Backlinko found content over 3,000 words got 77.2% more referring domain links than content under 1000 words, which is less than double.
The lesson is the usual one: open the study before you repeat its number. Our guide to tracing a backlink statistic to its source shows how.
What Google and Semrush say about word count
Semrush's own 2025 answer on content length says: "Google doesn't rank content based on word count." It adds that "A 1000-word page can outrank a 3,000-word article if it better aligns with searchers' needs."
Google has made a related point about thin content. In its November 2022 office hours, Lizzi Sassman said "word count alone is not indicative of thin content."
How to size a post you want links for
A rule of thumb we use, built from the findings above:
- Do not publish a short post and expect links. In Semrush's data, posts of 300 to 900 words got far fewer backlinks than average-length ones.
- Look at what already ranks and gets linked. Open the pages that rank for your topic, and the pages your competitors earned links to. Their length tells you what depth readers expect. Then put a top-ranking post into a backlink tool and read the anchors and the text around each link, as Ahrefs suggests. That can show which section or fact earned the links.
- Add length only when it adds something linkable. Original data, a full method, a template, a chart or a clear definition gives other writers a reason to cite you. Extra paragraphs that repeat the point do not.
- Check the links you earn. After launch, see which pages picked up links and compare their length and format with the ones that did not.
When you want to see whether those links are still there months later, SerpCrest's Bulk DoFollow / NoFollow Checker checks a list of URLs for live, dofollow and indexed status. For a checklist of what makes a page linkable, see our post on link earning versus link building.
