For at least a decade, Google has told us to stop worrying about how many times a keyword appears in our content. They’ve said their algorithms and natural language processing have become better at understanding the intent of content, so keyword density is less important than in the early SEO days.
But if you ask most SEO experts, they’ll still want to see your keyword appear multiple times in every piece of content.
So is Google telling the truth—that the importance of keyword density died a decade ago? Or is there still an ideal frequency to fit a keyword into your post?
To gain a little clarity and move this discussion toward more current trends, we surveyed top-ranking content from five search terms. We reviewed both how often and where keywords were placed in each post. Then, we used our results to put common keyword assumptions to the test.
We learned that keyword frequency matters much less than we expected it to, but there are still scenarios in which keyword placement is important.
How we ran the experiment
The goal of this experiment was to test common assumptions about where and how often you need to place keywords in content to help it rank highly in search engine results.
How did we find common assumptions? We reviewed popular posts about keyword density. If several pieces offered the same suggestion, we tested that assumption in our experiment. We then reviewed content from the first page of Google searches for five different search terms. Each of the terms:
- had at least 500 searches per month,
- was a keyphrase (meaning it had more than one word like “SEO certification”), and
- involved a topic in our SEO industry.
Since we were concerned with text, we skipped YouTube links. We also didn’t include content produced by Google, assuming that doing so would have an unnatural advantage in Google search results. In total, that left us with 42 articles from five search engine results pages (SERPs) to pore over. The first thing we looked for in each post was keyword density—the number of times a keyword or keyphrase appeared on a webpage, expressed as a percentage of the total word count.
To calculate keyword density, divide the number of keywords on a page by the total number of words, and then multiply by 100. So if you’ve used a keyword 5 times in a 1,000-word post, your keyword density is 0.5%:
(5 keywords/1,000 total words) x 100 = .005, or 0.5%
Next, we looked at where keywords appeared in a piece of content. In the advice we read, the most commonly suggested locations for keyword placement were in the title and in major subheads (H2s).
We counted only exact keyphrase matches . So if a blog post had a sentence like, “ There are several certifications you can get to show your SEO expertise ,” we didn’t count it as placing the term “SEO certification.” This policy kept the experiment manageable and allowed us to gauge the importance of placing exact keyphrases.
Assumption #1: The ideal keyword density ranges from 0.5% to 2%
We saw different suggested keyword densities in the posts we read, but they all fell within the range of 0.5% to 2%. Our results showed that front-page content rarely hit that mark.
On average, keyword density across all 42 posts we reviewed was only 0.25% .
Just seven of the 42 reached 0.5%, and none got close to 2%.
Let’s put that into perspective. In a 1,000 word post with a keyword density of 2%, you’d see the keyword 20 times. In comparison, you’d see the keyword only 2.5 times for every 1,000 words of the articles in our review. If you're trying to imagine how long 1,000 words is, it gets to just halfway through "Assumption #2" in this article.
Why so low? In part, because 12 of the posts in our study didn’t use the exact keyphrase at all . Yet they landed a spot on page one for the search term. With that said, every post in the number one spot did use the exact keyphrase at least once somewhere in the body copy.
Takeaway: Stress search intent, not keyword density
Instead of looking for ways to fit a keyword into your post, focus on answering the intent of the keyword.
Our results showed that you don’t need to include a keyword often, or even at all, to land at the top of a SERP for a particular search term. But you do need to consider the intent a searcher has when they enter that keyword in the search bar.
Here are some of the front-page results from the search term “SEO certification.”
The intent behind that search term is obvious: What certifications are available, and are they worthwhile? If you write a post about how to create an SEO certification, you’ll have trouble getting it to rank, no matter how many times you use the keyword.
This is true because Google now uses natural language processing (NLP) to understand the content it crawls. With NLP, a search engine can make complicated and deductive assumptions about a piece of text. Google knows what your content is about, with or without matching keywords. The better your content matches search intent, the more Google wants it to be on page one.
Assumption #2: You need the exact keyphrase in your title (ideally at the beginning)
We found this assumption to be partly false across the top-ranking posts in our survey.
36% of the posts we reviewed didn’t use the exact matching keyphrase in their title.
Additionally, only 28% of them had the keyword at the beginning of the title.
We did, however, see some alternative version of the multiword keyphrases in 40 out of 42 of the titles. For example, the post titled “Crafting a Good Page Title for SEO” held the top spot for the keyphrase “SEO title.”
Takeaway: Be creative with keyword use in titles
While you shouldn’t abandon adding keywords to your titles altogether, you don’t have to be a stickler in how or where you place them. Keyphrases don’t necessarily need to be at the beginning, and top-ranking post titles often have keyword variations.
That’s great news, especially if you’re targeting long-tail keywords . Trying to fit a term like “quality management software for medical devices” into the ideal title limit of 60 characters is tough.
Don’t be afraid to break up those awkward keyphrases or use variations; many others are likely doing it. For example, most of the results for the medical device keyword use the abbreviation QMS to shorten things up.
Assumption #3: The ideal keyword should appear in at least one H2
It was a common suggestion in the content we read that placing a keyword in a main header—like the “Assumption #3” H2 above this paragraph—helps Google understand your post. The data says that’s probably true.
Just over half (55%) of the posts in our survey used a keyword in at least one H2.
While that’s not conclusive, it’s at least a strong enough trend that we can’t dismiss this assumption entirely.
Does it help to have keywords in multiple H2s? Not likely. Only 28% of the posts we reviewed did. Also, only one of the five posts that held a number one spot had keywords in more than one of their H2s.Using H2s to break up long sections of text does seem to be important, as every single post on our list did so.
Takeaway: Use H2s to make your post scannable (and include a keyword if possible)
Generally speaking, what’s good for readers is usually good for SEO. That rule definitely rings true when it comes to headers.
Our results don’t show that using H2s will automatically boost your SERP placement. But we do know that a logical progression of subheadings that match your title will make your content more scannable and digestible for readers. Here’s what we mean by a logical progression of H2s that match your title.
Let’s say you’re writing an article called “Why Synthetic Materials are Better for Workout Clothes.” Each H2 in the article should directly answer the question “Why?”
So you want H2’s like this:
- Synthetics breathe better than cotton
- Synthetics don’t hold moisture
- Synthetics are more durable
You can imagine the word “because” before each H2. “(Because) synthetics breath better...”
Now, let’s say you used these H2s instead:
- Running is becoming more popular
- The best workout clothes
- Stretching after exercise is important
These topics might be mentioned in your piece, but the H2s don’t follow in-line with the title. Organize your post with a logical progression of subheads. Then, include a keyword in one of them, as long as it doesn’t look forced.
Assumption #4: The ideal keyword density is the same for all keywords
Across the five keyphrases in our experiment, the averages for keyword density differed greatly.
For example, the average keyword density of the top posts for “SEO title” was 0.35%. Compare that to 0.12% for “Google ads tutorial,” the lowest of any group of posts we reviewed. That’s a 300% differential. Clearly, average keyword density changes depending on the keyword you’re targeting.
One really interesting trend emerged here. All longer keyphrases were used less often in posts than shorter, two-word keyphrases. It’s a micro-trend, given our small study. But it suggests that the longer the keyphrase, the less you need to use it to compete on SERPs. And tying into the point in Assumption #2, the longer keyphrase gives way to more variations.
Takeaway: Test existing content to find a keyword density benchmark
There’s no one benchmark for keyword density that applies to every keyword you’ll target. Instead, you’ll need to review existing high-ranking content to see what density level is appropriate and use that as your guide.
There are several SEO tools you can use to check existing content. We used Internet Marketing Ninja’s free Keyword Density Analysis Tool in our experiment. You just enter a post’s URL and the tool provides an instant report showing total words, keyword density, and more. This analysis takes time to complete, so use it for the most critical search terms in your strategy.
If you don’t have time to review content, there’s a good rule of thumb you can follow. The longer the keyword, the less important it is to squeeze an exact match into your post.
Assumption #5: You won’t win on SERPs by stuffing keywords
Of all the assumptions we read, we were most confident that this one was true. And it was. We found no evidence to suggest that keyword stuffing would help with ranking. In their post on the topic, Google defines keyword stuffing as “the practice of loading a webpage with keywords or numbers in an attempt to manipulate a site’s ranking in Google search results.”
The highest single keyword density we found in any of the 42 posts we surveyed was 1.26%.
If keyword stuffing worked, you can bet top-ranking posts would have a much higher density.
Excessive keyword placement can even harm your SEO. In the same post, Google very plainly says that keyword stuffing will “harm your site’s ranking.”
Takeaway: Use keywords for relevance, not ranking
When you use a keyword in your post, make sure it’s because it will help a reader understand what your post is about. Once you start placing keywords in extra spots to help your SEO, you’re in danger of a Google penalty.
Of course, there are times when frequent keyword usage happens naturally. To avoid triggering a Google penalty, use a tool like Yoast , which warns you if you’ve gone overboard on keyword usage. You can also review other top content to see what the density benchmark is. If your piece is stuffed with the term, consider switching out some of your keywords for synonyms.
Mostly, it's about readability and sounding natural.
Readers have picked up on this and can tell that stuffed articles don't tend to correlate with being authoritative. They learned to sniff them out after too much use in early years of SEO. Repetitive use of keywords is an easy signal that the article isn't there to provide value. Google has been reiterating this point for a while. Write authoritative content in a natural, readable manner .
Keyword practices are continually evolving
The most important lesson we learned from this experiment? You can’t rely on yesterday’s assumptions if you want to rank on today’s SERPs. Instead, you’ll need to keep up with evolving SEO trends , pay attention to Google’s algorithm updates , and, most of all, keep track of how your own content is trending for your targeted keywords.