Keyword density: the 2-3% rule Google never wrote
Old SEO guides, and plenty of agency reports, still say to aim for a keyword density of 2-3%. Google has never published that number. No percentage appears anywhere in Google's search documentation. What the docs do define is the failure mode: keyword stuffing, which Google's spam policies describe as “filling a web page with keywords or numbers in an attempt to manipulate rankings in Google Search results” (Google's spam policies). There is no ratio to hit. There is covering the topic the way a knowledgeable human would, in the words your buyers actually type.
Where the 2-3% number came from
The number is not a Google rule. It is a fossil. In the late 1990s, keyword density really was a ranking input: early engines leaned on raw term counts, and a page that said the phrase more times could genuinely outrank a page that said it less (the metric's own history). Practitioners hunted for a sweet spot, numbers like 2-3% hardened into repeatable advice, and keyword density checkers turned the advice into a score, because a percentage looks like science in a monthly report. The engines it was tuned for no longer exist. The metric survived because it was sellable, not because it was true.
The same era produced two siblings we debunk separately: the meta description character limit and the exact match domain. And like NAP consistency, there is a kernel of something real underneath, inflated into a number somebody can bill you for.
What Google's documentation actually says
Google's spam policies list examples of keyword stuffing, and one of them is “repeating the same words or phrases so often that it sounds unnatural” (spam policies). Note the direction: the only place repetition shows up in Google's documentation is as a violation. Google's own SEO starter guide says it plainly: “Excessively repeating the same words over and over (even in variations) is tiring for users, and keyword stuffing is against Google's spam policies” (SEO starter guide). The phrase “keyword density” does not appear in that guide at all. We checked.
Google's people have said the quiet part out loud for over a decade. John Mueller, Google's Search Advocate, told site owners in a 2014 office-hours hangout that keyword density is not something he would focus on, because search engines moved on from counting repetitions years ago (recorded hangout). Asked on Reddit whether keyword density was still an SEO factor in 2021, his answer was one word: no (via Search Engine Roundtable).
Even earlier, Matt Cutts, who ran Google's webspam team, recorded a 2011 video answer on ideal keyword density (What is the ideal keyword density of a page?): the first mentions help Google understand what the page is about, after that the benefit shrinks fast, and continuing to repeat the phrase puts you in keyword stuffing territory.
Chasing a density costs you twice
Write for a ratio and you pay on both sides of the transaction. On Google's side, copy engineered to repeat a phrase is the documented definition of the one thing the spam policies exist to demote. On the buyer's side, it reads exactly like what it is. A homeowner who lands on “our roof repair team offers roof repair for all your roof repair needs” does not think this company is the roof repair authority. They think the page is fake, and they hit the back button. That is a visitor you already earned, walking back to a competitor whose page just talked to them like a person.
Modern search is semantic. Synonyms count
The deeper reason the metric is dead: Google stopped counting words the way density assumes long ago. Google's own explanation of ranking describes a synonym system built to “find relevant documents even if they don't contain the exact words you used” (How Search Works). Related terms, synonyms, and the entities on the page all tell Google what you are about. A page about water heater replacement that mentions tank sizes, permits, and venting reads as more expert than one that repeats “water heater replacement” fourteen times, to the algorithm and to the customer.
What to do instead of counting
- Put the phrase where humans expect it. The title tag, the one H1, and the first paragraph. That is signal, not stuffing. One clear H1, not five competing ones.
- Use the words buyers actually type. If customers say “fix my AC”, do not write only “HVAC remediation services”. Their language, once, beats your jargon, ten times.
- Cover the topic like the expert you are. Answer the questions that come up on real jobs: cost, timeline, what can go wrong. Depth of coverage is what modern systems reward, not repetition. Word count is a cousin of this same myth, and we debunk it in is there an ideal word count.
- Read it aloud. If a sentence sounds like it was written to hit a number, it was, and both Google's docs and your customers treat it accordingly.
Frequently asked questions
Is keyword density a Google ranking factor?
No. Google's documentation never names a density, and Mueller has said directly that it is not something to focus on. The only repetition rule in the docs is the spam one: stuff the page and you match the documented definition of keyword stuffing.
What is the ideal keyword density for SEO?
There is no ideal percentage, from Google or anyone who can prove it. The 2-3% figure is a leftover from late-1990s engines that counted terms. Place the keyword where a human expects it, then write for the reader.
Can using a keyword too many times hurt my rankings?
Yes. “Repeating the same words or phrases so often that it sounds unnatural” is Google's own example of keyword stuffing. And before any algorithm reacts, the buyer already has: robotic copy loses the lead you paid to get to the page.
My SEO tool gives me a keyword density score. Should I use it?
As a smoke alarm at most. A very high score can flag copy that reads stuffed. Climbing toward a target percentage optimizes for a metric Google does not use, at the expense of the human holding the credit card.