So... inside GOOGLE Analytics today...
 

So... inside google analytics today I found 2 queries that I hadn't ranked for previously, nor have I ever attempted to do so. The two queries are:

"www.sexynations.com"
and
"www.sexfoction.com"

It struck me as extremely odd that the only "similar" site suggested is mine.

I think I've got it figured out, but I'd like to see who's smart enough to figure it out and/or if I'm actually right.

sex(y) and 'tion(s)' make the domains similar

What did I win?

a link that may or may not have SEO value

http://en.wikipedia.org/wiki/Soundex

given that google uses SQL... this is likely to be similar to if not the algorithm they use for this type of SERP

phonetics? Interesting but I'd go with the much more obvious and simple "most number of letters in the right order to the target" idea

well that's why I'm posting this for discussion, but I think that

you might be on to something.

It would seem very processor intensive with so many alternatives for each word/syllable ..maybe it's just used as a last ditch effort after typos etc in those "few" cases when nothing else matches like your example.

Your theory doesn't satisfy anu query other than your own site.

For example:

related:www.google.com

related:www.booble.com

..and thousands others.


As far as I know, google show similar sites based on what content and in return common keyword both site rank for.

For example, if two sites rank for "adult search engine" .. they must be related/similar.

Does it make sense?

Similar addresses/urls, similar pages and related sites are three different animals as far as Google is concerned aren't they?

The results page actualy says

"However, there are results from websites with similar addresses:" :)

My bad. I misunderstood first post as if its about "related:" thing.

lol.

yeah vick that's the difference, they obviously don't use SoundEX or a similar algorithm in the natural SERPs because they don't need to.. but when they find ZERO results, then the grasp at straws