how to measure influence in social media networks

Photo credit: Emergent by Design

Sinan Aral’s session on ‘how to measure influence in social media networks’ covered a lot in a very short space of time and each sub topic, could have been a standalone topic in itself. First up, he dove into the problems at hand. He touched on the obvious, that your average early-twenty-something is prolific at making data. They’re likely to be checking facebook every half an hour, uploading to twitter every hour, reviewing in yelp a couple of times a week and posting to flickr once a day. He touched on what The Economist called ‘the data deluge’ – the sheer flood of data coming at us from all angles and the potential to really use this data for true behavioural understanding.

He touched on the fact that the act of buying products is an inherently social activity, but more often than not the actual transaction process is anything but and we don’t generally get to see this kind of data from our friends. The rise of social commerce and collaborative consumption has been more than documented recently. But, he argued that if we can understand how behaviours spread from person to person, we could actually use the force for good. So, use the underlying science of peer-to-peer networks to promote the good things like exercise, productivity and healthy living and curb things like smoking. Essentially, encouraging people to help their friends to change behaviour in a fundamental way. He also set up an argument that the likes of Klout (among others) aren’t doing a good job of measuring social influence. In fact, they’re doing quite a superficial job.

We then dove deeper into influence and we talked about Ashton Kutcher being largely seen as influential on twitter, yet he doesn’t actually influence much behaviour. He’s got a lot of followers and a big microphone, but those followers aren’t necessarily doing what he tells them to do – so in fact he’s not a true influencer. He then paused on what he calls ‘the reflection problem’. Where us humans cluster in network space and time because ‘everything is contagious’. But, is this because of peer influence, or an alternate explanation? One explanation could be due to homophily, where  ‘birds of a feather, flock together’ and Aristotle’s wise words, “(people) love those who love themselves”. Another could be confounding factors, the fact that people that are friends are exposed to the same stimuli. Basically, on the surface things can look like influence, but can probably be attributed to the three phenomenon.

He then went through some fairly solid examples to illustrate what he calls the ‘iPad effect’ – intense early adopters are actually exaggerated homophily and that people that are more similar influence each other more. I recommend going through these examples as he goes through what features spread contagion best and the merits of passive and active viral features. As, features that make it more likely for people to share among friends mean that you have a greater likelihood of adoption. There’s also a virtuous cycle between peer adoption and engagement and we all know little to no correlation between clicks and actual acquisition on the likes of facebook.

He also touched on ‘interruptive’ versus ‘native’ (e.g. video content) advertising and that the likes of facebook is more geared towards the latter. We had a look at a passion project of his called The Social Cure, which is aiming to get people in South Africa to get tested for HIV Aids. It’s built on the insight that there are more SIM cards than people in the country and they’re untilising this love of mobile to spread the word via SMS peer-to-peer to get people to take a test. Friends encouraging friends to do good.

We then got our g33k on and saw a working example of mining and visualising data using NodeXL and Gephi to get an influence graph together. Surprisingly simple and kinda fun!

This is just a very quick summary of the talk and I really recommend taking a look at the SlideShare to fully digest. But, here’s some thoughts to takeaway:

– Think about the microphone.

– Think about interruptive versus native in different contexts.

– Usefulness.

Let me know what you think of the examples and his POV.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s