Title tags are an interesting HTML element, while it was originally used to give a title to a webpage it is now used for many functions. It is used by the search engines as an on-page optimization factor, it appears at the top of the web browser, it is often displayed in the SERP for your page and more recently is the default title used when the page is shared on social media. In the SERPs, Google now displays the first 600 pixels of your title tag, depending on the characters – this is typically 60 – 70 characters. From this, many myths have begun to appear. The first is that if your keyword is beyond a certain number of characters it will not be taken into account by Google; or at least given much less weight. From this belief, entire hosts of tools have appeared that tell you if your title tag is “too long”. Even Google has an alert for this in the legacy version of their Google Search Console.
Google is obviously secretive about how its algorithm works and from that, we end up with hearsay, assumptions and myths. The perfect fodder for SEO testing.
The test I setup was straightforward. First, my hypothesis: Google only “counts” keywords that reside in the first 75 characters of a title tag – for on-page optimization purposes. I created a fake keyword. One for which there are no results when you search Google. I then developed 14 html pages and used the same content on all of them to eliminate variables. The first page was my control; it had only the keyword in the title tag. The second page title tag contained 25 characters and then the keyword. On the third page, I placed the keyword after 50 characters, fourth after 75 characters and fifth after 100 characters. I increased the number of leading characters by 25 up to 300 characters on the 13th page. I decided I should have a page without the keyword in the title tag so that was page #14. To my surprise, all the pages (except #14)ranked for the keyword.
In true myth-busting fashion, I decided to take it to the ridiculous extreme. I created a new test set with 12 pages and same hypothesis. Even though I saw Google reads well past 75 characters I kept the hypothesis the same because I wanted to replicate my first test and to be honest, I wasn’t sure what limit to use. The first two pages contained no keyword and only the keyword in the title tag respectively. After that, I had 10 pages, but in this test Iincremented the number of leading characters in the title tag by 100. By the time I got to my 12thpage, I had a title tag with 1,000 characters in front of the keyword. As a quick aside, a 1,000-character title tag is huge. Looking at it in the code gave me a good chuckle.
Nevertheless, back to the test. I was amazed; Google ranked ALL the pages for the keyword! So, I have two separate tests that reveal that Google will not only read the entire title tag but attempts to utilize all of the content for optimization purposes.
What I love about running tests are the other things you learn while running them. In both tests I observed that the page with only the keyword in the title tag always ranked the best. I had assumed that the order of the ranking results would roughly mirror the length of the title tag. Meaning, the title tags with the fewest characters would (in general) out rank the longest ones. What I found, in both cases, was that this was not what happened. To illustrate, here is the ranking order from the first test set (character count incremented by 25):
Page #1 (only KW)
Page #8 (150 + KW)
Page #2 (25 + KW)
Page #6 (125 + KW)
Page #7 (150 + KW)
Page #9 (200 + KW)
Page #13 (300 + KW)
Page #3 (50 + KW)
Page #11 (250 + KW)
Page #5 (100 + KW)
Here is the ranking order from the second test (character count incremented by 100):
Page #2 (only KW)
Page #7 (500 + KW)
Page #5 (300 + KW)
Page #11 (900 + KW)
Page #10 (800 + KW)
Page #12 (1,000 + KW)
Page #3 (100 + KW)
Page #8 (600 + KW)
Page #6 (400 + KW)
Page #9 (700 + KW)
Looking at these lists it seems relatively random to me, but…we are not guessing here, we are attempting to deal in precision science. I put this data into Excel and created a scatter chart, it still appeared random. I added a trend line and it did slope up to the right, which means there appeared to be a slight correlation between shorter title tags and better rankings. Not trusting my “sad naked caveman eyeball” (as Rick Sanchez would say) I took the numbers and ran a correlation coefficient in Excel. For test #1 the result was .4 and for test # 3 the result was .26 – a weak correlation in both cases.
So where does all this leave us? I think it is safe to say that there is no practical limit to the size of the title tag and you certainly do not need to worry about arbitrary character limits. I have no doubt the appearance of the page title will affect click through rates from the SERPs or social media so we still need to optimize for both SEO and CTR.
What I think we can take from this is if you have secondary or even tertiary keywords you want to include you can safely place them beyond the 60 character limit, it will still be factored into your page’s optimization. Personally, I am still going to endeavor to place the important keyword near the beginning of the title tag but I will also be sure to include additional keywords near the end. For local clients I would test adding additional geographic areas. This could also be useful for keywords with multiple spellings (i.e. donut vs doughnut) or even a word with common misspellings. I suspect there are many other cool ways these findings could be used. Please let me know if you can think of others.
The cool thing I like about testing is while it usually answers your questions it often raises more for future testing. I have come up with several more title tag tests, I look forward to sharing them with you all in the future.