Science of Social blog

Figuring Out the Relationship Puzzle

By MikeW

Figuring Out the Relationship Puzzle

by Lithium Guru ‎09-16-2010 01:11 PM - edited ‎09-15-2012 12:03 AM

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 at mich8elwu.



Are your relationships important to you? I think that most of you would answer “yes,” since most people value their interpersonal relationships and want to develop them and maintain them. But are relationships (e.g. with customers) important to business? As more businesses shift towards a service-centric and subscription-based economy, customer relationship will definitely become more important. Companies are realizing that it is the repeated purchase from loyal customers rather than the initial point of sales that is generating the most revenues. Unlike the traditional, product-centric, and sales-based economy, the point of sales is actually the beginning of a long term relationship with your customers.


Many companies understand this paradigm well and are making strides here. They start (or attempt) to engage their customers through various channels (both traditional and social media) with the goal to build stronger customer relationships. But what are they attempting to build? What is a relationship anyway? Have you ever wondered what your customer relationship management (CRM) system is actually managing? If you dig into the data in your CRM system, what kinds of relationship data will you find? In today’s article, I want to discuss the topic of relationship from a sociology perspective and try to understand it at a deeper level.


The Components of a Relationship

Relationship Puzzle Small.gifEveryone seems to know the concept of relationship, but no one seem to be able to define it precisely, or knows how to measure it. The definition of relationship from actually looks nothing like the definition from the Merriam-Webster’s dictionary. But when sociologists speak of relationship, they mean a tie or a connection between two entities (these entities may be people, companies, cities, or even nations).


One reason that I like the sociologists’ perspective is because not only do they make the distinction between the presence and absence of relationships (ties or connections), they also introduce the concept of tie strength. This allows us to talk about the strength of the relationship (i.e. weak ties vs. strong ties). Moreover, sociologists have gone further and developed ways to quantify the strength of a tie.


The notion of tie strength was first introduced in 1973 by Prof. Mark Granovetter in his seminal work: The Strength of Weak Ties. He identified four different components of tie strength.

  1. Time: amount of time spent together
  2. Intensity: emotional intensity and the sense of closeness
  3. Trust: intimacy or mutual confiding
  4. Reciprocity: amount of reciprocal services

Subsequently sociologists have identified numerous predictors of tie strength. These predictors are contextual variables that are correlated with the strength of a relationship, but they are not the actual the relationship itself. This is a subtle but important point, because some of the components (i.e. time and reciprocity) can be measured, whereas others (i.e. intensity and trust) are not well-suited for measurement. As with any quantity that cannot be measured directly, statisticians can still obtain an estimate of their relative value by forecasting them with predictors.


Consider the first component, time, as an example. The actual amount of time that was spent together comprises the relationship. It is a part of the relationship. But the duration of relationship is only a predictor of tie strength; it is not the relationship. Likewise, the frequency of encounter is another predictor of tie strength, but not the relationship itself.


In fact, duration and frequency are both poor predictors of tie strength. Frequency tends to over-estimate the tie strength between neighbors, co-workers, or other people who happen to meet regularly due to situational constraints. However, there may not be any strong emotional intensity or trust between them. On the other hand, duration tends to over-estimate the strength of relationships between relatives. Because we tend to know relatives for a long period of time (we probably trust them too), but we may not actually spend much time together to develop the relationship. See the subtle differences here?


Similarly, homophily (the similarity in education, occupation, social-economic status, political profile, religion, etc), multiplexity (the overlap of interests, topics, or social contexts), and emotional support (amount of advice given/received) are all reasonable predictors of tie strength. However, they do not constitute the relationship itself as Granovetter defined it.


In the Business of Relationship

So why and how is this important for business? If you care about your customers and your relationship with them, then this will be important:

  1. By deconstructing a relationship into four components we have a better chance at quantifying the strength of any relationship. This will enable us to rank the strength of relationships and identify weak ties that potentially need attention.
  2. By understanding why a tie is weak, we can focus on the problematic components when trying to restore the relationship. For example, should you focus on developing trust or spending more time together?
  3. Finally, if we can understand which components we can affect, then we can build relationships more effectively, whether it is interpersonal relationships or customer relationships.

Prof. Peter Marsden (now at Harvard Univ.) and Prof. Karen Campbell (now at Vanderbilt Univ.) were among the early pioneers who carried out a full-scale statistical model to measure the strength of ties in social networks. They’ve found that time and intensity are the strongest and most fundamental components among the four. That means they are also the most difficult to affect. This is because time is often a scarce resource, and it only contributes positively to the relationship if the desire to spend time together is mutual. Otherwise, spending more time together can actually hurt the relationship by turning the intensity component negative (i.e. turning love to hate). So as a business, if you want to build stronger customer relationships by spending more time with them, then you better understand when your customers actually want to spend time with you.


The intensity component is even more fundamental and more difficult to manipulate, because part of it is just genetics, over which we have no control. Although not impossible, it is generally difficult for businesses and brands to create a sense of closeness and strong emotional intensity with their customers. For example, brands are starting to leverage Green Business initiative to boost the intensity component of their customer relationships. This creates a sense of “we are in this together for a greater cause” and it will increase the emotional intensity of customers with your brand. Since this leads to a stronger customer relationship, the net result is increased customer loyalty and higher customer lifetime value (CLV). Despite this, the intensity component in customer relationships is still much weaker than those in interpersonal relationships. So if you are a brand, don’t try too hard on this component.


So where should companies focus their effort in building stronger customer relationships? The answer is building trust and increasing reciprocity. Trust and reciprocity are two components of relationships that companies can impact effectively. Clearly greater trust can lead to higher adoption rate and therefore greater CLV. With social media, companies nowadays have many ways to build trust. They can institute a blog or tweet program to enabling company-to-customer transparency, or launch a community to enabling customer-to-customer transparency. Whatever you do, just remember this: being authentic and transparent is the key to building trust between any entities.


Finally, increasing reciprocity is another easy way for brands to build strong relationships with their customers. Remember that reciprocity is defined to be the amount of reciprocal services. So it is not just you serving your customers. You should let your customers serve you too (This is a key point to remember). Let them help you by letting them help other customers and reward them properly. This will create a cycle of reciprocity can sustain itself. Aside from the added benefit of reducing support cost, implementing a co-creation strategy is one of is the most effective way to increase reciprocity between your brand and your customers.



So what have we learned today? I am hoping that we’ve learned that there are four components to a relationship: Time, intensity, trust, and reciprocity.

  1. Time and intensity are the most fundamental and most difficult to leverage by companies.
  2. The intensity component is usually much stronger in interpersonal relationships when compared to customer relationships.
  3. Trust and reciprocity are the two components that companies can leverage effectively for building a stronger customer relationship.
  4. Brands can build trust by being more transparent and authentic, and they can increase reciprocity by implementing co-creation strategies.

Alright, I hope this post offered you a different perspective and gave you a deeper understanding on relationship from a sociological perspective. By understanding the sociology of relationship, we can leverage it in a business setting by applying it to the analysis of customer relationships. This is the power of science: finding general theories that can be applied again and again to different situations (provided that the assumed conditions of the theory are satisfied).


Lastly, I’ve must confess that I’ve been pretty busy lately. I’ve skipped a week of blogging last week and reduced my twitter activity to about 10-15 minutes a day. I know, that is bad. I’m a social media scientist; I am supposed to eat and breathe social media. But I’m also an analytics engineer, and I am trying to wrap up some coding and bug fixing before I take my little vacation for this year. Yup, I will be going to Japan for two weeks starting next week. So I might be a bit slower when it comes to replying to your comments. But I will reply! So speak your thoughts. I always love to hear them.



by Martijn Linssen(anon) on ‎09-16-2010 02:55 PM

Fine post Michael, not regretting my premature RT ;-)

I like the four aspects, but I'm missing Free Will, which you introduce later

Weak ties: are you sure you need to "restore the relationship"? I have very weak ties with very good friends living a very great distance away. We had a very strong relationship once, and it doesn't need much to get going. Whenever we see eachother again, it's back to usual and we get very intimate very quickly

I think companies can spend oceans of time and acres of intensity on customers, by the way, and they traditionally have: but that was all one-way via direct mail or mass marketing. Run a P&R project, let the press roll, and everything that comes out can be just copied endlessly
However, the relationship need is different now: people appreciate a more personal approach. Well, for a personal approach, you need persons - and that's when time and intensity become scarce

One last thing: we spend time together a.o. on Twitter and via email in asynchronicity; I find that a very interesting subject because it means we can receive far more attention than we can handle (information overload). It'll lead to a blog post one of these days for me


The important question for me is: does Social kill the asynchronous part, and force on the synchronous part? If I look at Twitter, it does. But I'm getting more and more @replies and DM's that are waiting for me in the morning when I open up my Twitter client - from synchronous, Twitter is going to asynchronous (when used intensively)


Food for thought...


Enjoy your holiday in Japan! Enjoy the relationships there, and leave the ones here alone - they'll survive the two weeks

by Lithium Guru on ‎09-16-2010 04:11 PM

Hello Martijn,


Thank you for the rapid RT and as well as the rapid comment here.


I think that free will is a way for people to choose whether or not to create a tie or to develop the relationship, but it is not the relationship itself. I know that is a very subtle point. I knew it will create some confusion when I wrote this. But it is actually an important distinction. I've actually talked about the mutual desire to connect is a basic requirement for tie formation in one of my earlier posts on How Do People Become Connected?


There are many ingredients that will contribute to our choice of creating a tie and building the strength of the tie. Homophily (socio-economic similarity), multiplexity (interest and contextual similarity), etc will all have an effect. But they are not the actual relationship itself. They contribute to it and correlate with with the tie strength. If I choose to be your friend at this very moment but never spend time, develop any emotional intensity, build trust, and foster reciprocity, then our tie remains weak. So the act of choosing to connect or that choice, is not considered as an actually part of the relationship by Granovetter.


Yes, time and intensity certainly play a more important role in the modern day social media. But the amount of time spend with a brand or the intensity (closeness) of a relationship with brands is usually very weak compare to those with our friends and families. Consider how you will feel if one of your favorite brand vanishes from the earth vs one of your close friends vanishing. That should tell how much more intense interpersonal relationships is compare to customer-brand relationships.


Whether relationship need to be restored depends on what kind of relationship it is. I believe that for interpersonal relationships, as in your examples, the relationship just need to be maintain, because the relationship is of high enough intensity that you will recall it very quicky. However, for relationship with brands, maybe periodic restoration is necessary.


Notice that people living very far away can be strong ties too. Geographical distance is not part of the tie strength formula proposed by Granovetter. In fact, if you have very intense emotion (because you are such a good friend in the past), lot of trust, and much reciprocity previously, then your relationship will still be considered a strong tie, even if you are not spending much time together now. Remember that intensity is the strongest component.


Finally, IMHO, I don't think social kill asynchronicity. It is the fact that our communication platform is persistent and searchable that kill it. Social is just social, it can be synchronous or asynchronous (as it has been for hundreds of years before there is any computer).


Thanks for the comment and the wishes. I will definitely enjoy Japan.

But keep the comments coming. I will check it and reply periodically.


by Gautam Ghosh(anon) on ‎09-16-2010 09:25 PM

An insightful post as usual, Michael !


I think more and more "social media experts" should start reading your blog!

Any book plans in the offing? :-)


by Lithium Guru on ‎09-17-2010 08:29 AM

Hello Gautam,


Thank you for the very nice and affirmative comment.


I am learning about social media too. I just take a different angle, and use a more scientific and mathematical approach to understand social media. I feel that finding true principles underlying various observations can benefit the community much more. And of course, since I'm a train scientist, might as well use put that skill to good use.


As for writing a book... there has been discussion about it. But it seems that I have too many responsibilities at Lithium now and cannot really focus on writing. I can't really just take 6 month off to write a book and forget all the other things that I need to do. So I will just have to wait and see. Maybe someday when I have enough material on this blog, I will be eaiser to turn it into a book.


So will I write a book? In the short term probably not. In the medium-longish term, yes!


Thanks again for commenting.


by Occasional Commentator Iceman65 on ‎09-20-2010 04:48 AM

Thanks Mike for sharing valuable insights about the importance of the relationship for the customer lifetime value (CLV). Your blog posts always get my brain going on since I am a fan of the sociologists perspective on business!


I would love to hear more about your thoughts on how to translate stronger customer relationship (higher CLV) into the context of online communities and social shopping business.

by Lithium Guru ‎09-20-2010 04:58 AM - edited ‎09-20-2010 05:17 AM

Hello Iceman65,


Thanks for the comment.


Glad to hear that you are a fan of sociological perspective on business. I feel that there are a lot that we can learn from the social sciences (sociology, social anthropology, social psychology, behavioral economics, etc), especially when it comes to social media. In that case, the "social" is actuall much more important than the technology. So applying social science principle will definitely be a recurring theme in my blog.


As for translating stronger customer relationship to higher CLV, you can follow the logic here: stronger customer relationship --> greater customer loyalty --> lower churn and attrition rate --> longer customer lifetime --> higher CLV.


Hope this answers your question. Thanks again for taking the time to comment here.

by GlennR(anon) on ‎09-22-2010 05:02 PM

Thanks for writing this post and exposing me to the components of tie strength. It is extremely relevant to me. I blog about it here:


I'll be reading.



by Lithium Guru on ‎09-23-2010 07:41 AM

Hello Glenn,


Thank you for the reference and cross-linking in your post.


You are absolutely right, that Granovetter's theory is more gear towards interpersonal relationships. However, through social media customers are interacting with brands at a personal level. That is, customers are interaction with a person on the other end, who represents the brand/company. That was why I started to think, maybe I can apply Granovetter's model to brand-customer relationships, and if I do, what might come out of it? But I agree that it may not translate over directly to business relationships.


However, there are many other relationship models that focus on different aspects of social behavioral economics of customers or people in general. I would love to see what each one of them can contribute to the business world. Granovetter's model is just a blatantly obvious first step, because his model is so well known, simple and elegant among sociologist.


Anyway, thanks for the comment, and hope to see you around Lithosphere.