Since we started to work on mapping influence patterns, I have been wondering if we could find easy recognizable patterns in influence maps. If so, we could probably predict influence patterns and the secret of ROI optimisation would be eventually revealed to CMOs !!! Stimulating thought.
The recent history of science showed that behind apparently unpredictable phenomena, patterns could in fact be identified.
Further, similar patterns could be applied to domains as diverse as weather forecasting, traffic modelling or the evolution of populations: this is chaos theory.
So, could chaos theory explain patterns of influence on social media and resolve one of the biggest social media marketing enigmas for brands?
I wanted to dig into this. I however quickly had to face my own limitations: I know nothing really on the sophisticated mathematics behind the chaos theory.
So, as a first step, I decided to study graphs and maps and look for similarities that could confirm my intuition.
Now, honestly, if you know anything about math or physics, just ignore this post. It’s most probably embarrassing.
However, if you don’t care so much about maths, but are interested in influence, stay tuned: you might be interested in what follows.
Graphic representations of simple nonlinear formulas could produce such beautiful designs as the one above called “Mandelbrot set”. Could an influence map look similar?
Could we build a similarly simple theory of influence based on simple structures and relationships that could be modelled? My first step consisted in constructing a simple influence theory based on these two principles:
Influence is hierarchical : Social scoring companies such as Klout or Kred build a hierarchy among social media users: people with higher scores are deemed more influential than people with lower scores.
While I believe that these scores are useless because they are not contextualized, we do produce lists of individuals that we rank from 1 to 100 in specific, contextual topics.
Influence is an iterative process where sharing and engaging being the base mechanism for growing social influence.
Theories such as Malcom Gladwell’s “tipping point” suggest that viral content gets amplified from individual to another creating a series of iterations and reaching out to the masses.
That meant we could build a map of influence starting from our top influencers and building hierarchical relationships, representing engagement, to their “next level” influencers and so on….
Wow! Can we have a look at this please?
This is what I ended up with:
Could this be the shape of influence ? Could this be a realistic representation of the impact of individuals on their community down from the top influencer to smaller influencers ?
I decided to compare this to our 'real' influence maps, which aim to graphically represent top influencers on a topic, their closest contacts and the relationships between them based on real engagement data on Twitter: pretty cool stuff really!
Here is a sample of what they look like:
I looked across many maps but I never found any that really looked like my ideal influence map. My intuition said communities maps could be categorized in different types but none on these looked like my theoretical map of influence : I obviously got something wrong in my hypothesis.
Looking again through many Traackr maps, this is what I actually learned about the reality of social influence:
Influence is not a hierarchical relationship: it is reciprocal.
Influence goes in circles
It’s not a one-way path from the center to the periphery.
No one single person is the ultimate source of influence or content on a specific topic. Top influencers get influenced back by their community.
Influence is not soley proportional to social performance or popularity
Some key individuals with a low first-degree audience can have a high impact on a community because they have very strong relationships that help the community work better.
Social influence works imperfectly. It often fails.
Information or content can die out or be ignore amid the flow of information. You can have silos or subgroups who do not get exposed to specific contents.
Takeaways for marketers
No algorithm or big data miracle is likely to soon resolve the complexity of influence. After all, it is about gaining insights into your community and focusing on the right individuals for you.
It’s really about creating the right relationships and authentically interacting with individuals.
Will social influence one day be explained by chaos theory? Maybe… but this is still up for demonstration.