"Social influence" is like love or great art, it's hard to define, but we know it when we see it.
The thinkers at New York based social consultancy Lucule think they've taken a big step forward on the trail of measuring and understanding the social behaviors that actually influence users: the qualities of social influence that actually pass along ideas and opinions.
At SXSW last month in Austin they showed off their new measurement framework they've dubbed Pente. Pente challenges the notion of people who are "influencers." This of course means that a brand following the trend of chasing popular people online may be running a fool's errand.
Indexes like Klout and Kred that are trying to prop up that idea are desperately trying to retro-fit more and more sources of social data to keep, well, their cred intact. But according to Pente they have the problem exactly backwards: it's not the "influencer" who changes behavior, it's much more about the message--and the recipient.
This spring the Pente framework was run on thousands of social media messages from all of the major networks and blogs. Lucule says the Pente framework captures actual attribution of action across networks.
Pente measures five aspects of how an idea is passed along through social communication:
• Message type
• Message form
• Level of recipient engagement
Jure Klepic is one of the brain trust at Lucule that worked on Pente. A well-known blogger and social media practitioner, he's also long been a critic of the idea that popular "influencers" could be meaningfully defined and then identified. He believes the Pente model studies the major forces in social networks that have real influence on our opinions and behaviors.
"The Pénte is a social media-planning model. In other words, it provides a framework with which a brand can input each of the five factors and manipulate them to get the best or most efficient output," says Klepic.
"In this way, Pénte provides a conceptual framework and associated methodology for developing and testing social media activity with the goal of optimizing specific behavioral responses. As such, the Pénte model is widely applicable for brands, companies and organizations that seek to influence behavior."
Klepic and partner Debra Kaye walked out their first research with the Pente framework at their SXSW panel about social influence. They showed how they tested the Klout / Kred model of social influencers against what they found with their framework.
Here’s the results from their research about the relationship between an influencer’s “score” and that person’s actual influence on behavior:
Their statistical analysis indicates that the influence score accounted for only 3% of the variation in response. That is, 97% of the differences in behavioral response were driven by factors other than social influence.
Instead, consumers are more prone to react positively if a relevant message is received in a certain format (“News you Can Use,” for example) at certain times of the day when the level of engagement with a particular device is optimal (smart phones during the day, tablets in the evenings.)
We want to leave the idea of “sender” as influence," says Klepic. "The mere fact that someone sends a message is too simplistic in social media. Just cause a message goes out in the “ether” doesn’t mean it has any impact at all. It is just broadcast.”
"That is the reason for the model,” he continues. “There are many other factors that relate to see if a message that is sent is actually received, seized upon and integrated. So the idea that because a “sender” has a high influence score is influential is what the model proves against. Many other factors have greater importance."
The bottom line is that brands should craft any social messages they send around the context of the recipient, and not hang their hopes on the pseudo-science of influencer scores. That's more challenging, but it's more realistic given the vast, complex world of social networks.
For more information on the study, visit www.luculeconsulting.com