I’ve been collecting information on how much traffic Google send to me across my various domain names. I’ve found the results rather interesting and quite useful, and so I thought I’d share them. Note: This is based only on four data points. Skip to the ‘related work’ section at the bottom for some links to studies which are more involved, but which lack my deeply elegant wordsmithery and the soothing pixelated green header that puremango has become world famous for… (ahem)
As part of the adwords product, Google provides a Keyword Tool which among other things gives you an estimate of the number of searches being made on a particular term. First things first, when using Keyword Tool for domain research, make absolutely sure you untick “broad” and tick “exact” instead, otherwise Google will include searches which contain the keyword (“Phrase”) or synonyms and related keywords (“Broad”). We’re not interested in this, we’re interested in the “Exact” number of searches.
I use Keyword Tool for a number of tasks; for instance at work we sometimes debate about whether we should call our widget a flertlewurter or a grumpinghoffer. A quick search on Keyword Tool might reveal that flertlewurter is searched for far more often than grumpinghoffer, and so it’s probably sensible to call our widget that, as that’s what people seem to expect it to be called.
The other thing I use Keyword Tool for is when thinking about domain names. If you can get a domain name which has a lot of exact searches, you’re quite likely to be able to get a good portion of the traffic. It’s the only bit of SEO voodoo that I pay attention to; I have seen this strategy work very well with my own eyes. Of course there’s nothing to say that Google won’t update their algorithm to stop the domain being such a strong signal, so I try to only register domains that also sound and look good. I have to have *a* domain name, it might as well be the traffic-attracting GeekWallpapers.com if my site is about geek wallpaper. Which in fact it is.
Enough with the background.
So we’ve established that by registering a domain name which matches very closely what people are searching for on google, you can rise to the top of google for that term pretty quickly.
But what proportion of the figure reported by Keyword Tool can you expect to actually receive if you have a number one spot?
Here’s my data for the keywords I currently rank #1 for..
exact globals, actual traffic, percentage
1600, 620, 38%
1000, 340, 34%
210, 50, 23%
320, 97, 30%
So there you go. From my small experience, a number one position will give you an average of about 30% of the global exacts per month. You can use this information to help you decide between domain names, or to estimate the traffic your competitors are getting.
There have already been studies into click distribution on Google SERPs (search engine result pages). One at Cornell University found that 50% of clicks were on the #1 spot, while other studies place it as low as 18%. The latter uses a similar methodology to my much smaller more informal study. The disparity between the 18% figure and previous estimates is put down to rounding inaccuracies in Google’s reported search volume.
For me the bottom line remains that you can expect to capture 30% of the exact searches if you hit number one on Google. If it turns out to be only 18% then that’s no big deal – you’ll also get traffic for a host of peripheral phrases which should make up the difference. Either way, when thinking of buying a domain or targeting a keyword, I’ll use the 30% figure as my guide for the amount of traffic I can expect to receive.