How many page title characters does Google index?
SHORT ANSWER: AT LEAST 213
Last year we were ran a number of tests that established categorically (to our satisfaction at least) that Google was only indexing 84 characters of the page title. Beyond that it looked like any keywords were just ignored, and we were trimming our sails accordingly.
You can see why. A Google that indexes 5000 characters of a page title is a Google that allows for some injudicious page title tag keyword stuffing. While Google might place much more weight on keywords that appeared at the start of the title tag, long page titles give leeway for people to dangle long tail phrases off a short-tail target page with equity, rather than building a dedicated long tail page and trying to funnel equity through to it. That can make a big quality difference for users.
There’s been a lot of chatter about and testing of title indexation recently – especially following Rory’s post the other week where he mentioned the 84 character limit – and our finding is now that the length has leapt to a minimum 213 characters. We did a limited test on a couple of domains we reserve for such things and came up with a convoluted but natty page title that stretched to 213 characters.
And guess what? All 213 were happily indexed.
A few other side observations from this (admittedly limited test).
- A PR1 domain with the keyword in the page title *and* the content outranked the PR2 domain where the keyword was just in the page title – suggesting that page titles have to be backed by onpage content. Obviously there is no anchor text in play here – this is all onpage. As a note on this: the PR2 domain we use will rank for longtail terms even where there is better information, so the signal for keywords far down the page title must be fairly weak.
- A Twitter feed we also used actually ranked higher than one of the test domains for the test phrase. Whether this was a particular index or a personalised result (unlikely, in this circumstance) it demonstrates that even though Twitter feeds typically lack both content and links – by design - Google still treats Tweets as an authority. Personally, I think that’s batshit, but meh.
I guess this might stem from the generally low quality of long tail landing pages. In ecommerce in particular, long tail pages are generally generated from individual product SKUs without any unique content or equity from the site navigation. And I’m wildly speculating from a starting point of zero research, but the supposed drop in longtail traffic some people have seen following the MayDay update could be tied into this.
Page title strategy
To date, page title strategy for the long tail has been typically either:
- Dangle a lot of long tail content from a short-tail page with inherent equity, or…
- Build plenty of optimised long tail landing pages and pass equity through the architecture
In light of these preliminary tests, strategy 1 looks a better bet, with the additional refinement that you can now add more keywords to your title tags. If further research in the wild bears this out, a suggested page title formula might be:
First block – under 70 characters, headline keywords, branding and call to action, seen by real people
“Bear Snacks for Hungry Grizzlies – great deals from Bearsnacks Ltd”
Second block – second-tier keywords, heavy hints to Google
“Food for Bears, Bear Feed, Snacks for Bears”
Third block – long tail extras, possible matches
“My bear is hungry, what do I feed my bear, what do bears eat”
All that would have to be backed up by onpage content and good site architecture and it might ultimately prove more worthwhile for you to carry on with targeted long tail pages but if you’ve got a site with reasonable pagerank to hand then it might be time to start running a few tests of your own. We’d be interested to see if you learn anything in line with (or against) our own experiences.
Oh – and while we’re here, if there’s anything you’d like us to test then just leave a note and we’ll see if we can help.