The core idea in Mark Earls’ most recent book “I’ll have what she’s having” is that the unit of analysis for social behaviour should be the group not the individual. Collaborating with Anthropologists Alex Bentley and Michael O’Brien the book suggests if we want to understand how ideas and behaviour spread we should look at the wood not the trees. The book references game theory, network science and evolutionary biology to investigate how diffusion works – how ideas or behaviour spread in a network.
In markets with many similar options and many people to learn from (e.g. paperback books) regular patterns like long tail distribution often occur. These real world patterns can be modelled using simple rules, such as undirected copying:
- A pool of individuals
- Lots of options to choose from
- From one time interval to the next, most choose by copying others
- A small percentage of the pool (1-10%) invents something new as opposed to copying others
- This process repeats over a series of time intervals
Network simulations using undirected copying rules are interesting: a few ideas gain favour, their fortunes rise and they dominate. Others rise and quickly fall. Most gain no popularity. The dynamic is unpredictable. The seismic changes in mass behaviour that we often see in real life, where an idea or behaviour suddenly becomes popular from left field, readily emerge. Earls et al compare real world data of trends in baby names or buzzwords in academic papers to demonstrate how similar the real world and these modelled “cascades” are.
If their analysis is correct:
- We are less autonomous than we presume. Whilst it is a few cases of innovation or “individual learning” that create the spark, it is the large amounts of copying or “social learning” that follows that leads to ideas or behaviour changing.
- It is hard to predict which idea or behaviour will spread. Social behaviour is analogised to the weather, a complex dynamic system with emergent properties.
The implication is that we should understand if our category is governed by social learning (e.g. pop music) or individual learning (e.g. car insurance) by analysing the market distribution. Categories governed by social learning lead us to recognise what we can control, what we cannot, and adjust our strategy and tactics accordingly:
- Research the social context not just the product in isolation. Observation in situ will take in norms, social influences and actual behaviour. A questionnaire on the buying decision after the event may not.
- Help people copy each other. Think white ipod headphones: the tech is in the pocket but the wire is in plain sight.
- Create things people can share. My view is that strategically this is just old-fashioned impact: marketing’s applied creativity making people respond ‘that’s funny/clever/interesting’. Tactically – it can be achieved however you like – through any channel rather than necessarily being digital (viral, game, app). It may be about what’s already there – facilitating existing groups or activities.
- Spread the risk. Campaign planning of old (Gantt charts for the year ahead drafted in June) seem ill-equipped to cope. Trying a range of ideas on a small-scale then investing in the ones with most traction seems most sensible.
- Respond to events. An extension of the above: the fickle zeitgeist. Having ideas in the bank allows you flexibility. A seed you were planning to sow next spring might fare better today.
I enjoyed the book. It is full of ideas and engagingly written, however I had to work hard to tease out relevant implications. I can see three challenges to their ideas. First, when you change the unit of analysis from the individual to the collective it is harder to establish concrete conclusions and hypotheses. More data and real world examples will help here. Second, measurement. Currently we can only infer undirected copying from market buying patterns, after the event. Third, when you apply the idea of undirected copying to the human scale you meet resistance. Western society inculcates individuality without us noticing.
Go and ask anyone with a star tattoo.