Both Google and Bing have added many social search features over the past year. There’s also been talk about using “social signals” to help rank regular search results. But are either of the major search engines actually using those social signals to rank regular search results? A bit, they tell me. In particular, your stature on Twitter could help influence how a page ranks in web search.
I gave Bing and Google six questions about how they use social data from Twitter and Facebook. In particular, I wanted to know how that data influenced regular web search results, not the impact it has on the dedicated social search tools they have.
Social Search Ranking…
For example, both Google and Bing offer a way to see results that are written by or shared by your friends, search that’s directly influenced by people you know:
Both Google and Bing also offer a way to see content being shared in real time on the social networks:
The links above take you to our past coverage of these dedicated social search services, explaining how they gather content from social networks and leverage social signals from those networks to help decide what should rank well.
…Isn’t Web Search Ranking
What happens within those social search tools is completely different from what happens when you do an ordinary web search, where Google and Bing try to decide which pages to rank tops from the billions they have indexed from across the web.
For example, Google uses more than 200 different “signals” to decide how rank those pages, in response to any search. Some of these signals are well-known, such as:
- PageRank, how authoritative a page is deemed to be
- Anchor text pointing at a page
- HTML title tag, and whether the words you searched for appear within it
There are many other factors beyond those I’ve listed. Bing also uses a complex recipe — or algorithm — of signals to determine rankings.
What’s not clear is whether social signals have been included in this mix. For example, both search engines have ways to determine if someone seems to be an “authority” or a trusted figure on Twitter, which they use within their social search features. But for web search, if that person tweets a URL, does the URL get a boost because a human authority pointed people at it?
Web Page Authority Vs. Human Authority
That’s how it works when it comes to web pages. Some pages are deemed more trustworthy than others. If those pages link to other pages, then what they point at gains reputation in Google and Bing’s ranking systems.
To some degree, “humans” on the web have pages that already represent their authority. For example, my Twitter page has a Google PageRank score of 7 out of 10, which is an above average degree of authority in Google’s link counting world. Things I link to from that page — via my tweets — potentially get more credit than things someone whose Twitter page has a lower PageRank score.
(NOTE: PageRank scores for Twitter pages are much different if you’re logged in and may show higher scores. This seems to be a result of the new Twitter interface that has been introduced. I’ll be checking with Google and Twitter more about this, but I’d trust the “logged out” scores more).
PageRank, Meet SocialRank?
Time for some “Buts.” Twitter also uses what’s called a “nofollow” tag that prevents links that are tweeted from getting credit. Hang on to that thought.
The bigger “But” is that even though my page has a PageRank score, it might still be useful for Google (or Bing) to give me something like a “SocialRank” or “HumanRank” or “AuthorRank” score independent of that. This would be a way for them to know how much authority that people — rather than pages representing people — have on social networks, and to let those people have a signal that influences rankings.
Who You Are On Twitter Counts
Enough with the background. Are they doing it? Yes, at least for Twitter. Both Google and Bing tell me that who you are as a person on Twitter can impact how well a page does in regular web search. Authoritative people on Twitter lend their authority to pages they tweet.
When it comes to Facebook, Bing says it doesn’t try to calculate someone’s authority. Google says it does, in some limited cases. However, I’m double-checking on this, as I think that might not be correct.
No Nofollow For The Twitter Firehose
Remember that whole “nofollow” thing I mentioned earlier? This is a way for anyone to tag a link and effectively say to search engines, “Don’t count this link as a vote.”
Nofollow is commonly used across the web in places where services can’t vouch for the links that might be posted on them. Wikipedia uses it, because it has so many volunteer editors that it can’t trust them all. Many blog commenting systems use it. At Twitter, links it publishes on the web are tagged with nofollow.
So, while Bing and Google do have a human signal they can assess from Twitter, supposedly they have no link signal that they can also count. But as it turns out, both of them get what’s called the “firehose” of data from Twitter. This is a constant stream of what people are tweeting.
In that firehose, links do not carry nofollow attributes — so there is some link credit that counts, in some cases. Bing tells me:
We take into consideration how often a link has been tweeted or retweeted, as well as the authority of the Twitter users that shared the link.
Google tells me:
We use the data only in limited situations, not for all of general websearch.
Questions & Responses
Below are the questions I sent over to both services to compile this story, along with their responses. Bing’s are paraphrased from a phone interview I had with them; Google’s are from the email replies I received.
1) If an article is retweeted or referenced much in Twitter, do you count that as a signal outside of finding any non-nofollowed links that may naturally result from it?
We do look at the social authority of a user. We look at how many people you follow, how many follow you, and this can add a little weight to a listing in regular search results. It carries much more weight in Bing Social Search, where tweets from more authoritative people will flow to the top when best match relevancy is used.
Yes, we do use it as a signal. It is used as a signal in our organic and news rankings. We also use it to enhance our news universal by marking how many people shared an article [NOTE: see the end of this article for more about that].
2) Do you try to calculate the authority of someone who tweets that might be assigned to their Twitter page. Do you try to “know,” if you will, who they are?
Yes. We do calculate the authority of someone who tweets. For known public figures or publishers, we do associate them with who they are. (For example, query for Danny Sullivan)
Yes we do compute and use author quality. We don’t know who anyone is in real life :-)
3) Do you calculate whether a link should carry more weight depending on the person who tweets it?
Yes we do use this as a signal, especially in the “Top links” section [of Google Realtime Search]. Author authority is independent of PageRank, but it is currently only used in limited situations in ordinary web search.
4) Do you track links shared within Facebook, either through personal walls or fan pages?
Yes. We look at links shared that are marked as “Everyone,” and links shared from Facebook fan pages.
We treat links shared on Facebook fan pages the same as we treat tweeted links. We have no personal wall data from Facebook.
5) Do you try to calculate the authority of someone on Facebook, either say via their personal wall or their fan page.
We don’t do this on Facebook. On Facebook, we only get what’s public, only updates and things you’ve posted to everyone as viewable. We don’t get things only shared with friends, so we don’t know how authoritative you are on Facebook. There isn’t the whole convenient retweet mechanism we see on Twitter.
We do see valuable content shared by Facebook users, even though we only get what’s public. For example when Gary Coleman died we saw a video from Different Strokes, saying his favorite line “what ya talk’in ’bout Willis” gain popularity. It happened to be what a lot of people are sharing on the day he passed away.
Again, the treatment is the same as for Twitter. And we have no personal wall data from Facebook.
6) Do you calculate whether a link should carry more weight depending on the person who shared it on Facebook?
We can tell if something is of quality on Facbook by leveraging Twitter. If the same link is shared in both places, it’s more likely to be legitimate.
Same as question 5.
7) And just to be really clear, the new Facebook data is not yet being used in ordinary web search, right? (asked only of Bing, because it was only relevant to them)
Social Data For Display Use Vs. Rankings
That last question leads me to an issue about using social data for display purposes, rather than ranking purposes. It’s another important distinction to note.
In The Wake Of Bing & Facebook, Google Web Search Tests Getting More Social covers how Google is showing “Shared By” figures next to news stories. This shows how many people on Twitter are tweeting about a particular news item.
The news stories aren’t being ranked according to the number of shares. It’s simply additional information being displayed, more on an FYI basis than anything else.
Similarly, Bing has said in the near future, results it lists will also show the number of “Facebook Likes” next to them, in some situations. Again, this is simply a display usage, an FYI for the searcher. The results themselves haven’t been ranked by number of likes, nor is that part of the overall ranking signals.
Yet. Stay tuned, because over time, it is likely that social signals will gain more weight in search ranking systems, I’d suspect.
TwitterRank & Retweets As The New Link Building
In the end, it’s clear that Twitter data especially plays a role in web search, these days. Who you are is being understood. Are you a trusted authority or not? If there’s PageRank for pages, both search engines have a form of TwitterRank for people.
Meanwhile, retweets serve as a new form a link building. Get your page mentioned in tweets by authoritative people, and that can help your ranking in regular search results, to a degree.