wheel - 8:50 pm on Dec 8, 2012 (gmt 0)
Apologies if this is a silly question, but what "social signals" is Google actually able to see?
That's not a silly question. It's THE question. And the answer is, not a whole bunch. What's social media today? Twitter and Facebook right? Everyone's all humped up about them.
You think Google decided to incorporate crawls (because IIRC, they don't get the twitter feed anymore) from TWO websites run by commercial entities bound to be their competitors, into their algo? Just how stupid do you think these people are?
And if they didn't incorporate data from those two sites, then I'd say their social signals are pretty weak.
And if they're not measuring that, then what exactly is it that you think they're measuring? The amount of bullcrap thrown around about 'social media' signals is astonishing. Nobody really defines what it is they claim to be measuring, and then asks themselves if it actually is likely or realistic to be measuring those signals. Likes on facebook? C'mon, puleeze. The fact is, I haven't seen or heard of anyone who actually can say 'they are measuring this, if I increase this specific thing, my rankings increase'. It's always just vague 'social signals'.
Why is that? Wait, I got it:
Not to mention, there is good money to be made doing the whole social bit as part of a service to offer.
And now we know what's driving all this social media hookum. SEO companies preying on the corporate clueless. No different than when it was 'all about links'.
I don't actually know if they are doing it, but this:
Yep. Activity - identifiable, traceable, recordable activity on a site is the common denominator.
seems like something they can measure, directly, without depending on someone else. But we still don't know how noisy that signal is. And I can provide lots of counter examples. So I think it's still doubtful. I'd provide wiki as an example - that site has to be sending high bounce rate signals all day long. In and out in 30 seconds, never a second page visited or a link clicked. Yet there's the ranking. (so at a minimum, that suggested that *if* visitor behavior is measured, it's only as a boost, and wouldn't seem to act as a negative, i.e. the crappiest bounce rate won't hurt, it just won't help.)