Social influence is the opposite of the weather. Unlike the weather, everyone is talking about social influence–and it seems like everyone is doing something about it.
Companies are springing up like ragweed everywhere in the “social influence” field. Some, like Radian6, are trying to listen and discern social influence. Klout and Kred are trying to score social influence of individuals. Everywhere we have companies trying to sell social influence through social networks—social SEO, F-Commerce, LinkedIn strategies—from agencies, consultants, marketing automation companies, and on and on.
It’s easy to understand why so many people are making a mad dash into the field of social influence. It’s clear that social influence exists, and equally clear it has a profound effect on how each of us views the marketplace, on what we buy and on what we recommend to others. It also has a growing influence on the brand equity of companies. Overnight companies like Pinterest suddenly appear, fully grown with enormous market capitalization.
Now a mythology of social influence is rising. It’s not a “myth” in the sense something isn’t true. It’s that there’s a mythology growing about this magical dust called social influence. The magic dust exists, we can see it in action and it’s very real. It’s not clear how it works, and how we create it--even though we have growing legends of those who have succeeded.
It’s natural that businesses want to find out how to use social influence to win in the marketplace, that’s easy enough to understand. But for all the serious effort put into understanding how to influence customers socially, the problem is staggeringly complex.
Maggie Fox also has a smart post called, "The Evolution of Social Media Measurement."
Sinan Aral of NYU’s Operations and Management Services Department, writes today as well about how effectively social influence can be used to create “contagions” of viral influence, borrowing from epidemiology.
Aral points out the logical minefield in market research of separating causation from correlation. For example, I may buy a North Face jacket because I see a lot of people wearing them who I think are cool. Or I may buy one without that influence, but just because I like the look of it. The first case is causation—others cause me to make the purchase. The second is correlation—we all buy North Face because we have similar tastes, but none of us cause any of the others to buy it. Correlation does not mean causation has happened. The problem is that in social media correlation is easy to measure, and causation is difficult.
Why is correlation is a big problem in measuring social influence? On the one hand we can see more and more people signing up for Pinterest. But who or what is responsible for influencing people to do so? Is it me? You? Did Pinterest get key influencers like TechCrunch to try it out? We can see the correlation, but we’ll be damned if we can accurately track the causation reliably.
Aral says: “we know little about how to create cascades of positive social behavior.”
And there’s the rub. Why is it hard? It's because there are dozens of factors that pull and tug at us to create the social influence we each feel. This is difficult to dissect, but here’s a top 10 list of how social influence is created by a message sent across a social network:
Major flows of social influence are easy to see. As I write this Tim Tebow is being traded from the Denver Broncos of the NFL to the New York Jets, and once again he’s a hot topic on Twitter. The number one reason I’m interested is the content of the message—I’m interested in where Tim Tebow will play next season and like a lot of NFL geeks I follow the off season trades and college draft way too closely.
Bitly published research last Fall on the “half life” of social network messages, based on measuring how many people passed along a given bit.ly link. Here’s the pass-along effect of 1,000 messages from five different social networks (including “direct,” which includes links shared through e-mail, SMS, IM, etc.)
Lifespan of messages across social networks. Time is measured in 2 hour increments. (from Bit.ly)
There’s no research readily available that measures the decay time of people seeing an average tweet, probably because of how often people check their feed, the average number of followers and the average number of following are all over the place.
But I’ll push on and say that the Bitly research suggests that there are an enormous number of your followers who never see an overwhelming number of your tweets. If a Tweet falls in the woods and no is around to read it does it have any social influence at all?
(I do want to put in a plug for one of the best books on the subject, Phillip Sheldrake’s The Business of Influence in which he covers the field of other research efforts on measuring social influence, calls out the various companies trying to do so, and then suggests a balanced scorecard for social influence measurement.)
This is the very real complexity of measuring and then doing anything about social influence. We social media practitioners trying to understand how social influence works are like wildcatters drilling for oil. We look at rock formations, maps of other claims, and we drill, hoping to find oil. Often we don’t. Occasionally we do, but it’s a small well. Often we see a few who have found a geyser and staked their claim to a big oil field.
But understanding social influence is new and the problem as vast as human behavior. Give respect to the complexity of the problem.
Rohn Jay Miller is Managing Partner of Content & Social, an agency in Minneapolis that helps traditional marketing organizations adopt content and social strategies, working with their senior management to bring positive change to how they do business.