Because word count is easy to measure and quality is not. A number in the corner of the editor feels like progress. Telling a writer to hit 2,000 words is simple. Telling them to answer the question better than anyone else already has is much harder, so the number wins by default.
We get asked this on nearly every content project. Someone has read that long form content ranks better, and they want to know the target. The honest answer is that there is no target, and chasing one is how good pages turn into padded ones.
This is a debate worth settling properly, because the wrong answer costs real money. Teams pay writers by the word for filler nobody reads. So let us go through what Google actually says, what the data actually shows, and what we do instead.
There is no ideal word count. The right length is however many words it takes to answer the question completely and then stop. A comparison of two platforms might need 2,000 words. A definition of a technical term might need 300. Both can rank if they fully satisfy the search behind them.
This frustrates people who want a number, so here is a more useful way to think about it. Word count is an output of doing the work, not an input you set beforehand. You research a topic, you find you have seven real things to say about it, and the length falls out of that.
The moment you flip it around and set the length first, you have to find words to fill the gap. Those words come from somewhere: restating the introduction, defining terms your reader already knows, adding a history section nobody asked for. Every one of them makes the page worse.
No, and Google says so directly. Its guidance on creating helpful content asks writers to check themselves with this question: "Are you writing to a particular word count because you've heard or read that Google has a preferred word count?" Google answers its own question in the same line: "(No, we don't.)"
That parenthetical has been sitting in Google's documentation for years, and it is about as unambiguous as search guidance ever gets. There is no threshold you cross. There is no penalty for being short. A 400 word page that nails the answer is not disadvantaged against a 2,400 word page that circles it.
The same document draws a harder line on why padding backfires. It flags content that is "primarily made to attract visits from search engines" as a problem, and warns that using automation "to produce content for the primary purpose of manipulating search rankings" violates Google's spam policies. Writing to a quota sits uncomfortably close to that description.
Because thorough pages are usually longer, not because longer pages are usually better. Backlinko's analysis of 11.8 million Google search results found that "The average word count of a Google top 10 result is 1,447 words". That average is real. The causal story people attach to it is not.
Look at what that number is actually measuring. Pages that rank in the top ten tend to cover a topic properly, cite sources, address the follow up questions, and get linked to. All of those behaviours produce words as a side effect. The length is a symptom of the effort, not the cause of the ranking.
The same study is blunt about the correlation. It reports no direct relationship between word count and position, and suggests long form content may help a page reach the first page while doing nothing to move it up once it is there. Adding 800 words to a page that already ranks fifth will not make it rank first.
That study found something far more actionable in the same dataset: "The #1 result in Google has an average of 3.8x more backlinks than positions #2-#10". If you are choosing between writing another thousand words and earning one good link, the data points at the link.
Not in the way people hope. Google's documentation on its AI features is explicit that "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." There is no length threshold that gets you cited, and no minimum that gets you excluded.
What Google does require is ordinary and specific. To show up as a supporting link, "a page must be indexed and eligible to be shown in Google Search with a snippet, fulfilling the Search technical requirements". That is a technical bar, not a content length bar. If your page is not indexed or blocks snippets, no word count saves it.
Here is where we do think length interacts with AI search, though, and it is the opposite of what most people assume. Answer engines pull passages, not pages. A model looking for a clean answer to a specific question wants a tight, self contained paragraph it can lift. Burying that paragraph inside 2,500 words of preamble makes it harder to find, not easier. We wrote more about how this plays out in practice in our guide to optimising for Google AI Overviews.
When the search behind it has a short answer. If someone types a definition query, a conversion question, or a yes or no comparison, they want the answer in the first screen. Padding that page out to hit a word target actively hurts it, because the reader has to dig for what they came for.
Transactional and navigational pages are the clearest case. A pricing page, a contact page, and a service page all convert on clarity rather than depth. We have never once improved a pricing page by making it longer, and we have improved plenty by cutting them in half.
Getting this right starts with reading the query properly rather than guessing at it. Our breakdown of search intent covers how to tell what someone actually wants before you decide how much to write, which is the step most content briefs skip entirely.
When the topic has real depth and the reader has real questions. Technical guides, platform comparisons, migration walkthroughs, and anything where the reader is making an expensive decision all justify length. In those cases people want completeness, and stopping early leaves them going elsewhere to finish.
The test we use is simple. Every section has to survive the question "would a reader be worse off if we deleted this?" If the answer is no, it goes. A 2,200 word article where every section passes that test is strong. A 2,200 word article that passes it in eight sections and fails it in five should be 1,400 words.
Length also earns its keep when it lets you cover more real questions on one page. A guide that answers nine genuine follow up questions is more useful, and more citable, than three thin pages answering three each. The words are doing work there. That is the difference.
Look at behaviour, not at the word counter. Thin content shows up as pages that get impressions but almost no clicks, or clicks with immediate returns to search. Those are readers arriving, scanning, and deciding you did not answer them. That is a real problem, and it is not the same as being short.
Google Search Console is where this becomes visible. Filter to the queries a page ranks for, then read the page as if you had typed each one. If a query brings up a question the page never addresses, that is a genuine gap. Fill that gap. The page will get longer as a consequence, which is fine, because the length came from the gap rather than from a target.
The failure mode we see most often is the opposite diagnosis. A page underperforms, someone assumes it is too short, and 900 generic words get bolted on. Traffic does not move, because the missing thing was never words. It was usually a mismatch between what the page says and what the searcher wanted.
Thin is really a question about substance rather than length. The useful test is whether your page adds anything the pages already ranking do not, which is the argument we make in our post on what information gain is and how you add it to your content.
Only if you have found a specific unanswered question. Adding words for their own sake is the most common wasted effort in content SEO. Adding a section that answers something readers are clearly searching for, and that your page currently ignores, is one of the highest return edits available.
The distinction matters because updating an existing page is usually a better use of an hour than writing a new one. The page already has history, links, and rankings to build on. We went through when to update versus when to start fresh in our piece on content refreshes versus new posts.
When we do expand a page, we track what changed and why. If the new section came from a real query in Search Console, we expect movement within a few weeks. If nothing moves, the section was probably filler dressed up as research, and we take it back out.
Stop setting word count targets in your content briefs and start setting question targets. Write the list of questions the piece must answer, research each one properly, and let the length be whatever it turns out to be. Then cut every paragraph that does not answer something. That single change usually improves quality and reduces cost at the same time.
If you want a number to hold onto, use it as a floor for diagnosis rather than a target for writing. A page under 400 words on a competitive topic is worth reviewing. A page at 1,900 words is not automatically good. Neither number tells you anything on its own.
We take this position strongly because we have watched teams spend serious budget on words that made their sites worse. If you are staring at a content plan full of word count targets and wondering whether it is working, we are happy to look at it with you and say honestly what we would change. You can find us at phoenix.studio.
Tell us where you want to go. We'll tell you how we'd get you there.