Showing results for 
Show  only  | Search instead for 
Did you mean: 

Google+ vs. Facebook vs. Twitter vs. ... Part 1: The Strength of Bidirectional Connections

Lithium Alumni (Retired)

Dr Michael WuMichael Wu, Ph.D. is 927iC9C1FD6224627807Lithium's Principal Scientist of Analytics, digging into the complex dynamics of social interaction and group behavior in online communities and social networks.


Michael was voted a 2010 Influential Leader by CRM Magazine for his work on predictive social analytics and its application to Social CRM.He's a regular blogger on the Lithosphere's Building Community blog and previously wrote in the Analytic Science blog. You can follow him on Twitter or Google+.



Since its launch in late June, Google+ has been a hot topic of conversation. So before I proceed with either the engagement or the gamification series, I think it is appropriate to take a short detour to discuss Google+.


I think by now everyone has most likely had the chance to play with Google+ (or at least know what it is). If you still don’t know, don’t miss the boat. There are already 25+ million people signed up. I am not going to repeat Google’s pitch, so take a look at the demo from Google. What I’m going to talk about today, are some of the network properties and social principles behind some popular social platforms. Of course, I will compare and contrast them to the new kids on the block (i.e. Google+), and talk about their behavioral implications.



name facebook_web.pngFacebook certainly brings great utility to its users by helping them maintain the relationships in their lives. However, people’s relationships in the real world are very complex. Not every relationship is the same, and every friend is unique in some way (see Social Network Analysis 101). By collapsing all the different relationships into one bucket, called friends, Facebook created the problem of conflicting social spheres. Some groups just don’t mix in real life, and some information is not intended for all friends, not even among close friends.


Humans deal with this problem by spatial-temporal segregation. We simply meet different groups of people at different time and different places. But on Facebook, we are stuck; we either don’t share, or share to all our friends. Although Facebook has subsequently implemented features that allow their users to put their friends into groups, these features are band-aid fixes that didn’t drive much adoption and are therefore ineffective.


This is precisely the problem that Google+ is trying to address with Circles. I am certain that social researcher like Paul Adams, an ex-Googler now at Facebook, has infused much social insight and contributed to the present success of Google+. Unfortunately, he is not allowed to comment on the Google+ project due to his non-disclosure agreement. Before I say more about Google+, let me point out some interesting and distinctive network properties of Facebook.


Bidirectionality: A Double Edged Sword

handshake_sxc191184_8572_web.jpgFriendships on Facebook are bidirectional. That means both parties need to consent in order connect. Anyone can initiate the connection, but the other person must accept it to create a tie. This type of bidirectional, mutual, and reciprocating relationship is very important, because it keeps the relationship and information content more relevant. As a result, the signal-to-noise ratio on Facebook is fairly high compared to other networks that only require unidirectional consent, such as Twitter. The same can be said about Linkedin. Because Linkedin requires bidirectional consent to connect, the information content on Linkedin tend to be more relevant compare to Twitter.


However, bidirectional consent is not without a cost. One of the drawbacks is that it requires the coordination and alignment of intent between two people. This is harder to achieve than it seems, and it limits the growth rate of the network. Why? Because it takes more time for people to build their network. You can easily spend the weekend and follow 1,000 people on Twitter, but it’s much harder to get 1,000 friends on Facebook. Even if you can, it’s going to take you much longer to build a network of same size on Facebook due to its bidirectional consent.


Ironically, this offered Facebook a strong upper hand against Google+. The simple explanation (the psychological perspective) is that, because it takes more time and effort for people to build their network on Facebook, they are less likely to forfeit all their investment and start over again. This creates a psychological barrier preventing Facebook members from switching. Although Google+ may gain more new members, people will tend to cling on to their Facebook network, purely because of the amount of time they already invested in it.


A more rigorous explanation involves the network effect. Although Facebook is still 30 times bigger than Google+ now, its true strength is not the sheer volume of its user base, but the cohesiveness of the networks. Facebook contains many strong ties: families, close friends, high school buddies, etc. Having a lot of these strong connections is what made Facebook extremely sticky. Strictly from a network value perspective, the more strongly interconnected a person is on Facebook, the more utility they can derive from the network (i.e. network effect). Conversely, there is more they have to lose by switching to Google+, where these connections may not be present on the new network. Therefore, users who have many friends and can bring many friends to the Google+ are precisely those that are hardest for to get, because they are least likely to switch. Users who have little or no friends are the easiest to switch, yet they are least effective for driving mass adoption of the new network.



Although Google+ has a phenomenal growth rate, the strength of bidirectional connectivity on Facebook shouldn’t be overlooked. Despite the fact that bidirectionally-connected networks have a slower growth potential, Facebook offers:

  1. More relevant relationship: stronger ties that makes the network very cohesive and sticky
  2. More relevant content: higher signal-to-noise ratio (SNR)


Conference FTW-badge_web.pngSince the past couple of posts I wrote had fairly low traction and almost no kudos, I’m going to do something different in the next couple of weeks to see if things change. I will split the longer articles into short ones and post them more frequently. So later this week, let’s see what we can learn from the Twitter network.


By the way, I am in Philadelphia now. I’ve been invited to participate in a Wharton Gamification Symposium at the Mack Center for Technological Innovation. This symposium starts today and it’s organized by Wharton Prof. Kevin Werbach and NY Law School Prof. Dan Hunter. They managed to gather a very distinguished crowd of thought leaders, academic researchers, scientists, practitioners, entrepreneurs, investors, as well as policy makers. I’m very honored to be invited. Since there are so many experts at this symposium, I believe it will be an interesting learning experience for everyone. So I will have much to share after I return. In the mean time, we can talk about Facebook and Google+. Follow me on G+.



Not applicable

Isn't the weekness of the Facebook model that it doesn't support interest graphs? Those are more valuable for certain purposes. The problem is it takes too much maintenance to create the graphs using circles.


On the getting love, always a problem for bloggers, take heart: do you have twitter account? And direct twitter retweet and facebook like links might be more effective. Putting the latest articles first might help.

Lithium Alumni (Retired)

Great post, Michael.  


Bi-directional ties are by definition stronger than unidirectional ties, but it seems to me that whether you want weak or strong ties depends on your objectives.  For example, weak ties are better for problem-solving and innovation, because your strong ties are more likely to know the same things that you know, and what you need for those purposes are people who knows things you *don't* know. 


I've noticed that I personally have added more of my professional relationships to Google+, because Circles allow me to better manage them separately from family and personal friends.  But maybe the support  Google+ provides for unidirectional ties is another reason why Google+ is a better place for professional networks.  Perhaps, like LinkedIn,  G+ fiills a niche rather than taking on FB in a zero-sum game. 

Hi Mike!

What about risk from spam with friendship in G+ - who adds me to cicles without tie establishing?

Google seems made small input threshold for new users - but it can cause inflation of new and quickly ties?

Not applicable

When Facebook got twitter envy they introduced the "ignore" option on the friend request. So if someone sent me a request and I clicked on "ignore" the person would still be able to see my public posts on their newsfeed.. effectively making a "following" relationship in a bi-directional network

Lithium Alumni (Retired)

Hello ToddH,


Thank you for the comment. I’ve been at the Gamification Symposium at Wharton all day today. So I apologize for the delay response.


It is true that one of the weakness of FB is that it doesn’t support interest graph. But that problem is precisely the problem that we cannot group people with the same interest into groups. Like I said, technically FB does allow you to do that. However it faces the low adoption issues that make it practically ineffective. BTW, an interest graph is really just a circle in Google+ with the same interest. Mathematically circles and network connections can have the similar interpretation, because they are topologically the same.


The problem of time and effort required to maintain the interest graph using circle will be discuss in the 3rd post of this series. So I encourage you to come back next week for that. I apologize that I had to make you wait due to the experimentation I did with my own posting habit. Being a scientist, I always like to experiment with new ways to improve my reader’s experience. I tend to not use FB for work/research related stuff, but I do share my posts on Twitter and Linkedin.


BTW, this is the latest article. Because my articles goes into great depth, it is often quite long. This article is nearly 3500 words, which I’ve split up into 4 posts. I plan to post them every Mon and Thur.


Anyway, thank again for the comment.



Lithium Alumni (Retired)

Hello JoeC,


Thank you for the comment.


You are totally right. Strong ties and weak ties have different function and utilities. In fact, the value of strong ties vs. weak ties have been discussed in an earlier post I’ve written (see From Weak Ties to Strong Ties). But in general, the value of strong ties are in the relationship itself, where are the value of weak ties are in their number and diversity, since you can have so many of them.


Compare to FB, G+ is definitely have a much better niche in managing professional relationship than do personal relationships, and Circles on G+ definitely allow you to more efficiently manage your relationships. Whether G+ is taking on FB or not, due to the attention economy and the fact that we only have 24 hours a day, it will affect the amount of time people spend on FB. It may try to fill a gap. But because our time and attention is already so limited, by filling a gap, it might actually end up hurting FB’s traffic.


Anyway. Great observation and thanks for the comment. Thanks again for taking the time to comment.



Lithium Alumni (Retired)

Hello Andrei,


Thank you for bring up this question. It is a question that I will address in the 3rd installment of this mini-series.


In brief, so I don’t give out the whole series now, unidirectionally connected networks are definitely more prone to spam. This is one of the reason that Google+ try enforce the real name policy. Basically G+ is trying to ensure that you are really who you are. Still, that doesn’t prevent people from following a 10,000 users over a weekend, the only thing that is stopping people from doing that is probably due to the negative network effect (i.e. network congestion). Basically if they try to follow 10,000 people, their stream will be so  noisy that it would be not usable for the user. So the noise level also serves as a cap on how many people can actually follow.


Anyway, glad to see you back again. I’m glad we fixed that FB connect problem. Hope to see you next time.


Lithium Alumni (Retired)

Hello Gautam,


Thank you for the comment and pointing out the ignore feature on FB.


In my next post, we will discuss Twitter, and this point you mentioned is very relevant there. For now, I just like to point out that bidirectional network and unidirectional networks are not mutually exclusive. As you said, you can have following, which is a unidirectional connection on FB. Likewise, on Unidirectionally-connected networks, such as Twitter, you can also have reciprocal connection too. It has to do with the ability and the consent mechanism for people to connect. That is, whether you ask for people’s permission to connect or not.


Thanks again for the comment. Stay tuned for the next post on Twitter! See you next time.