Thought Leaders in Big Data: Dmitri Williams, Professor USC and CEO of Ninja Metrics (Part 3)

Dmitri Williams: We had a model which started explaining these social influence ripple effects, and catch what we call social whales. These are people who may or may not have a lot of behavior on their own but clearly are causing the behavior of others. That’s where the commercial stuff starts to become much more interesting and obvious. If you know who the largest influencers are in a social system, this is extremely useful on a business model side for acquisition, retention, and for monetization.

It doesn’t matter if you’re buying movie tickets, playing video games, or consuming streaming content online. The business is always interested in finding the right people and getting them and all their friends to spend more. There’s one use case for each of those three sets when you know who the social whale is. You have to get more of them, find out where they came from, and keep the ones who are most important. You might not have realized who they were and piggyback on them and their relationships.

The social whales ARE different from just knowing the big spenders, which is what most marketing and most thinking is based on. That doesn’t explain all behavior. In fact, the behavior that we measure, we call it social value. It’s the percentage of behavior, which is caused by other people. That could account for anywhere between 5% to 80% of behavior in a system. The average is probably more – around 10% to 20% depending on what sector or market you’re talking about. You’re talking about significant chunks of behavior which have never been able to be touched or measured before.

Sramana Mitra: Let’s take the example of the social whale. In marketing terminology, we call them influencers, right?

Dmitri Williams: Yes.

Sramana Mitra: You said on Twitter, people are talking about what they think. When you look at that influencer behavior on Twitter, there is quite a significant and fairly easy way to figure out who the big influencers are in a particular sector because they have huge followings, they tweet regularly, and influence the thought processes and behavior of their community. This is happening all over the Internet right now and maybe categories are shaped by this influencer behavior. Can you help us understand what behavioral data offers that is significantly above and beyond that?

Dmitri Williams: It’s a simple one word answer. It’s proof. When you think about basing influence on tweeting, the behavior you’re measuring is someone else tweeting. I say something and you listen or you follow. From that, we assume that you then take an action. If I get on and I say, “I love Cheerios.” Then you say, “Dmitri says this is great. I’m going to retweet this. I’m going to start talking about Cheerios too.” What we have there is a decent measure of celebrity status. It’s more at the brand level, but it isn’t anywhere close to purchasing Cheerios.

This segment is part 3 in the series : Thought Leaders in Big Data: Dmitri Williams, Professor USC and CEO of Ninja Metrics
1 2 3 4 5 6 7