From Links to Votes: Search and Social Hawaii
Ian Lurie (@PortentInt), founder of Portent Interactive in Seattle is up in this session. He is going to be talking about social media and votes and signals for ranking. For more, check out Ian’s presentation slides.
He starts with a story.
Clients call you up and tell you that you’re a genius because of fantastic rankings or whatever it is. You can’t figure it out. Ian admits that he is the Eeyore of the search industry, so he’s not initially optimistic about it.
He is showing research where the two sites with the most authority were not No. 1 or 2 in rankings for a common keyword, “jeans”. And even the on-page grade was low. Most of the pages aren’t optimized; most of them are images.
Then he looked at brand mentions. Some of them have a lot of mentions, but the No. 3 spot does not have many mentions. He keeps digging. Then he looks at Facebook Likes. He does not see the correlation between the Likes and the rankings either. Some of the pages with more than a million Likes aren’t in the top rankings.
Then he pulled the average number of likes and comments per post for the brands. This is public data, including data that Google and Bing will get. The Facebook performance of the different brands really starts to weigh in. The level of engagement is showing as a factor.
Twitter doesn’t matter as much. But low posts and friends may hurt you. He is showing that Guess Jeans has an imbalance of friends to followers and low posts.
This is not a hard formula, but there is a correlation. The quality of the community matters to Google.
On-page = Supreme Court
Links = President
Quality = Congress
Social media = The voters
You need social media and you need it badly. The data he just showed is that the performance of Facebook and Twitter is every bit as important as the links you would have obtained a few years ago in SEO.
Nofollow links seem like they matter. The easiest way to connect a vote to a brand is a link. They are not going to award authority to the nofollow link but they will trace the link back as a form of citation. Other things Ian wants to look at is G+ but the sample is too small at this point. It’s all marketers at this point.
The fact that Google says it’s a ranking signal scares the bejesus out of him.
The takeaway: If you are not doing social media, you have your head in your rear end. Even if you are a pure SEO, you have to do it.
He has an acronym as a final thought for social:
Curate, Reward, Answer, Prove
3 Replies to “From Links to Votes: Search and Social Hawaii”
I’m butting in here to clarify, because my comment during the talk wasn’t totally clear. Nofollowed links appear to pass some kind of authority wherever they appear – on web sites, in social media, etc. – based on the context. That fits with the rational surfer model, too: Links in places where no sane person would click them pass nothing. Links in obvious, readily-clicked locations pass more authority.
What no one knows is HOW MUCH authority nofollow links pass, or under exactly what circumstances. The one thing that’s a safe bet: The search engines use links in tweets and Facebook posts to know what site the tweet/post is referencing. Then they use that to assign authority.
My best proof of this was the blog post that I had ranking top-3 for ‘agency tips’ within 2 hours of publication. There is absolutely no other explanation. The page had zero link authority, yet it beat 20-30 other high-authority sites. No other way Google found the article – they saw a small tweet storm around it, followed the link, and did the math.
hope this helps,
You confused me a little bit in the article just below that where you mentioned that Nofollow links seem like they matter. You didn’t mention the existence of any other nofollow links other than the social media ones like Twitter.
Where you said that nofollow links matter because they are counted as citations, were you speaking of Twitter specifically or do you mean nofollow links in general? If you are speaking of nofollow links in general, do he give you any other specific cases of analysis to back that up or is this his first reasonably conclusive case that made him realize it was happening?
The reason I ask is because Matt Cutts admitted that Tweets are used as a signal and at least for indexing purposes. I never saw him mention if there were any other situations of nofollow links being counted in any way. It would be very interesting data indeed if someone definitively determined that all nofollow links were counted as valuable citations.