Social networks have many dimensions. One dimension consists of the relationships we have to other people. Another consists of the things we show interest in through our activities on these networks. The first dimension is the basis of the social graph; the second forms the interest graph. Rene Reinsberg provides a simple definition: “Essentially, the Interest Graph is an online representation of individuals’ interests, with people and interests being the nodes of the graph.”

People are more than the sum of their connections to other people. They’re not just interacting with each other when they tweet or update their Facebook status, they’re interacting with ideas, issues, tastes—all the things they care about.

Our networks are full of people that don’t share our interests. I could care less about pet costumes, for instance, and yet, I still see that content in my Facebook feed because some of my friends do. This effect is particularly problematic on Facebook, where freshness is given more weight than most variables in deciding which content to serve up and feature.

Jonathan Sundqvist makes the case that Facebook is essentially a social graph, whereas Twitter is closer to an interest graph:

On Facebook you probably know all the people already. You care about what they think and how they feel. All the “friends“ you make on Facebook are mutually reciprocated. In other words there is a social bond between you and that other person. However weak or strong that may be.

Twitter on the other hand is formed around the interest graph. You start to follow people because you think they say interesting things or are interesting in other ways. You don’t necessarily care how they are doing on a personal level but take great interest in what they know on a professional level…

But what promise does the interest graph hold to consumers and companies? Consumers stand to gain an online experience that is exactly what they need it to be, crafted to their tastes without an extensive learning curve or manual calibration phase. They will discover more of what they’re after, be that media (Pandora and Netflix have led the way here), product recommendations from people like them, or even new friends that share their interests. Companies, simply put, have the opportunity to provide this rich experience to consumers, to learn far more about them than previously possible at this scale, to connect them to more of the things they care about, and to those that share their interests and passions.

How might all this play out once we’re really tapping into the interest graph’s potential? Let’s say I’m shopping for a new HDTV for gaming, and I post a request for recommendations to Twitter. What information would I most like to see in the replies? Ideally, it would be specific product recommendations from those like me, whether they’re strangers or not.  By “interest mining” my Twitter stream, a company could learn which games I have tweeted about most, and by looking at the percentage of my tweets that are about gaming, they could classify me as a “casual gamer.” They could then look for information about my space requirements, and see that I’ve tweeted about living in an apartment. Having established this profile, they can now tell me what people with similar profiles have been recommending. It could look something like this:

Ian, casual gamers that play 1st person shooters and have space limitations have been tweeting about X HDTV: [link to product page]

I click, and I’m taken to custom product page, where profile-matched content from Twitter, product reviews and Facebook is presented. I connect my Facebook account, and in a few seconds, I’m seeing recommendations from people that are even more like me. It’s all dynamic, sortable, filterable, and presented in a seamless, navigable user experience. Imagine being the retailer or HDTV manufacturer that can provide me with this page, without me first filling out an online profile—or even directly asking for it.

This is just one possibility, but the future of the interest graph is so much more than product recommendations. It has the power to create richer profiles of brand advocates, as Brian Solis found when analyzing @Starbuck’s twitter followers (hint: coffee is actually one of the weakest links between them). It can help us plan our careers by showing us the skills shared by those with the jobs we aspire to.  It can inject a new layer of relevancy into the search experience by boosting results in which others like us found value, as expressed by onsite behavior and other factors. And because our interests are constantly in flux, it can also provide the “elasticity” that is the basis of some interesting apps like Color.

The social graph isn’t going anywhere, but soon it will be understood that it provides an incomplete picture of who we really are. For brands, innovating around the interest graph is ultimately a way to provide even more value to consumers. Learn more about them to deliver more of what they want.

3 Responses to “Shared fascinations: The future of the interest graph and interest mining”

  1. Don’t get me wrong, it’s totally important. But I’d also like to see a less exclusive focus on this dimension. Thanks for stopping by.  for stopping by. 

  2. I think the data that we get from social graphs is highly important. We just need to understand how to actually use the information

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>