We have been testing social signals a ton. We have pushed in almost every single way imagined:
Likes
Tweets
Followers
Shares
Page Likes
Page Shares
+1
G+ Share
We even assumed AuthorRank was a factor, so we tested the same things with high powered authors. Guess what? It has no effect on rankings. SEO news blogs can slice data however they want and show correlation but it is not a cause for rankings.
We have tested with fun tools like SocialAdr, SYNND, buying accounts, etc. None are worth it unless you are preparing for it to count in the future. Chances are in that case you are trying to build a real brand, so do RCS to get those followers. Don’t do black hat tactics for these social signals.
SYNND had one effect on ranking, after months of hammering an entire site, two pages hit PR1. It was purely from well-spun bookmarks (we controlled it). It never actually ranked.
For Bing, you can get “trending” and a temporary boost in ranking when you add likes to a Page at a fast rate, or add a ton of shares. But there are better ways to rank in Bing.
The bottom line is that you should ignore social signals for now. We have some more advanced things we are testing regarding how PostRank (acquired by Google) focused data, but it is a very small part of the algorithm that should not distract you.
At most, for black hat, Google+ and Twitter are great for indexing tiers.
And a random fact is that Google puts a lot of weight on Diigo bookmarks. They’re awful.
One piece of data we will also include is that we have not been able to test Chrome bookmarks. Considering all the other factors being so weak, we don’t expect this one to take off either.
Just build links.