Social signals don't seem to be very important, yet.

[Sgt_Kickaxe]

This is just an observation, In looking at search results for an issue I'm trying to resolve I noticed that a lot of top ranked pages have no social stickiness while pages ranked 6th onward sometimes have hundreds of tweets and likes and pluses. In fact the top ranked pages don't even have social profiles a good percentage of the time, albeit on my very small scale searching this evening.

If social signals were becoming important I'd have expected to see the very authoritative pages ranked 6-7-8 to do a little better vs the top 3 that don't have a social mention between them. I know there are other factors involved but one could assume that if a page is well mentioned socially it also has natural incoming links, etc. The quality was sometimes better on the pages with social mention too.

Are you seeing any shift in importance from backlinks to social mentions on your site?

[anallawalla]

In fact the top ranked pages don't even have social profiles

Google might use social signals as a tie-breaker somewhere in the domain trust part of the algorithm. It might use it to add to an AuthorRank score, which won't apply to many business sites where there are no identifiable authors. I don't believe that every ranking situation requires the presence of social signals - in fact, if I were designing the algo, I might not use social signals at all in many niches.

But niches that use a lot of social media, such as celebrity content, can benefit from social signals.

[Sgt_Kickaxe]

That's true, anallawalla, it's not one algo fits all. In some ways I'd wager it's more like the stock market where stocks/sites are grouped into indexes and you've got your leaders in each individual sector. All of your metrics are weighed and measured over time by an algo much like stockbrokers analyze company performance results.

If the above analogy is accurate then Google will have perfected their keyword groupings much the same as the stock market indexes individual companies into their appropriate sector. Like the stock market, if your sector is faltering or gaining in popularity your site might be somewhat along for the ride. This system would need to take advertising into account so it's not a total numbers game, it's a metrics improvement game.

Hmmm, you've got me wondering which sites are in my sector. If I can find the sites that didn't advertise(I didn't) and that have relatively the same traffic and social metrics I would expect to see their traffic graph rise and fall at the same rates mine did this year.

Of course changes at Google, index adjustments and metrics tweaks would skew those graphs a bit it should be relatively easy to say "yup, that site is in my sector" anyway. If social signals become more important what will become less important?

[wheel]

But niches that use a lot of social media, such as celebrity content, can benefit from social signals.

A perfect example of vague feel-good statements from the SEO community.

Until folks are willing to define specifics that make sense and are measurable, I'll continue to call this an SEO salesjob. Social signals? WTF is that supposed to be?

I'd personally believe this hookum a lot more if the blackhats were running rampant using social media to rank, but AFAIK, they aint. Which once again, leads me to believe that all of this is being driven by SEO consultants selling stuff, rather than actual results.

[CainIV]

A correlation study at SEOMoz shown a statistically significant connection between social sharing and mentions and rankings across a range of niches.

That being said, the study also appropriately noted that it was the chicken or egg phenomenon - were the sites more socially shared because they were inherently higher quality and more relevant, or were they socially shared, and because of that, ultimately received higher scores in search.

One thing is for sure - social media helps to spread high quality content both on the website and off, which can lead to better performance.

[chicagohh]

Social endorsements are laughably easy to fake. If they are done *right*, it's time consuming to figure them out by hand. It must be torture for a bot. Friends, likes, tweets, followers, endorsements, profiles... fake, fake, fake. I am sure Google uses social patterns to help define quality, but I can't imagine they are taking it too seriously.

Here is this years link farm and log file spam rolled into one tasty treat:

> add real links to a good article

> add a few fake comments on a new YT video
> link the video out to the target article
> give the article 40 or so real/fake FB likes
> seed in some primer comments
> link to the author's Twitter page
> which has 900 followers and only following 25
> Tweet a few times about the new video

The "real" site can be one part of the chain or even just a linked from the article.

[wheel]

A correlation study at SEOMoz shown a statistically significant connection between social sharing and mentions and rankings across a range of niches.

That being said, the study also appropriately noted that it was the chicken or egg phenomenon - were the sites more socially shared because they were inherently higher quality and more relevant, or were they socially shared, and because of that, ultimately received higher scores in search.

As you noted, correlation <> causation. And personally, when I look at it, I tend to believe that it's not causation. The rankings are causing the social media presence, not the converse.

And with an organization as strong on SEO as SEOMoz, the fact that they left the correlation<> causation link unproven is a valid reason to be a conspiracy theorist. if they could've proven social media could cause higher rankings, I'd really expect that they would've. But they didn't.

One thing is for sure - social media helps to spread high quality content both on the website and off, which can lead to better performance.

Actually, that's not for sure at all. In fact, other than in exceptions, I don't agree at all. It breaks almost every 'rule' that I know of for most business' marketing plans. People aren't going to social media to research purposes or purchase purposes, and people aren't buying as the result of social media recommendations. Yeah, there are exceptions, but that's all they are.

far better to promote and advertise online where people are actively seeking and researching your products and services, rather than some vague garbage dumpster of general content.

If you're going to spend time doing promoting outside the search engines, targetted and relevant spots are going to serve far better results. and outside of that, I'd imagine stuff like targetted magazine ads or radio spots will still to better than social media.

[wheel]

And just an FYI. I have a reasonable social media presence, entirely as the result of other marketing that I do - I get tweets and stuff. And if we remove the ranking boost (which I believe doesn't exist in any measurable fashion) then we get to this:

One thing is for sure - social media helps to spread high quality content both on the website and off, which can lead to better performance.

And my experience has been that for me, it drives exactly 0 business. In fact all my non-google pushes drive almost 0 business. My mainstream media mentions in national daily newspapers, radio interviews, book sales, all drive almost 0 new business. I do it for other reasons, if I was measuring $ return, I wouldn't be doing it. People think that a mention in a national paper is going to drive sales. It doesn't.

I'm suspicious that there may be another ranking factor that Google's using to muddy the waters, but I don't believe it's social media. It may be some trait that they're using to measure 'brands', or type in traffic, or some wierd quality of links. There's stuff that's ranking that I don't know why they're ranking - but it's not social signals IMO.

[mrguy]

And my experience has been that for me, it drives exactly 0 business. In fact all my non-google pushes drive almost 0 business. My mainstream media mentions in national daily newspapers, radio interviews, book sales, all drive almost 0 new business. I do it for other reasons, if I was measuring $ return, I wouldn't be doing it. People think that a mention in a national paper is going to drive sales. It doesn't.

I've had the opposite on a few businesses I work with. The very things you mention drove a whole of traffic to them that was non existent prior to that.

I think whether or not social or other methods will work depends entirely on the business you're in.