By Guy Champniss


I still have vivid memories of when I was little and I started to misbehave, my mother would bend down and whisper something in my ear. Each time, it was the same thing. And each time, it stopped me dead in my tracks. ‘People are watching you’ she’d say. I’d look around and not be able to see anyone taking a blind bit of notice, but it was still enough to get me to pack it in; that somehow making the social context of my (mis)behaviour salient was sufficient to change my behaviour.

Fast forward four decades, and the broad field of behavioural economics has shone a brilliant light on the myriad ways our behaviours deviate from what normative models would tell us, on how seemingly irrelevant forces influence our behaviour and on how nudges can start, stop or twist our heuristics to move us toward other behavioural responses. And behavioural economics continues to grow in popularity with marketers and advertisers, with agencies seeing the field as a rich source of insight to better understand their customers and consumers (thanks in no small part to the efforts of Rory Sutherland and Nick Southgate at the IPA, and Alain Samson with the BE Guide and this site). This post is definitely aimed at those in the marketing field and tries to explore an obvious question: how does social context fit within this new behavioural economics-driven view of customers and marketing? Put another way, how does social context influence the way we – as customers – respond and behave when it comes to buying stuff?

There has been plenty of work that looks at the use of norms (descriptive and injunctive) to influence our decisions (e.g. filling out tax returns honestly and on time, or hanging up our towels in hotel rooms to save laundry efforts), but I’d like to argue that it’s still pretty early days in terms of looking more closely at how social context changes our behaviours; social context doesn’t just steer us toward different behaviours, but can have a fundamental influence on how we want to behave. This is less about an external social influence and more about some internal state – what we can call a social identity influence.

This is potentially very important for marketers and others who engage consumers on a day to day basis, since having a better understanding of both the force and speed of these internal social eddy currents could at best make all the difference between marketing success and disaster, and at worst, help explain why sometimes we as customers seem just so fickle. The reality is more likely we’re not this fickle – just more complex and susceptible to the social machinery that whirrs around us.

Social identity theory is certainly not new. But its widespread recognition within marketing is.

The theory proposes that our self-concept is made up of both a personal identity and then an array of social identities, which are selected depending on the social situation we find ourselves in. Social identities are linked to groups of similar individuals, and are selected when that particular group may help us understand how to act and succeed in a given situation. Put another way, social identities help us remove social ambiguity (no-one likes ambiguity).

Social identities are also important because they carry with them specific behaviours that support the identity, which means there are also behaviours that would betray the social identity. So marketers should be careful not to try and request a behaviour from their consumers when a specific social identity is primed that eschews that behaviour. This was very likely the challenge faced by Unilever with their Axe deodorant advertising in Italy. The brand was busy spending large amounts of money on TV advertising, only to realise their target market (young men) were frequently sitting at home watching these ads next to their mother on the sofa. At that moment in time, the behaviour encouraged by the ads (and for which Axe is famous, albeit in a tongue in cheek manner) clashed with the primed social identity of ‘dutiful son’. Nescafé at its launch made a similar mistake, encouraging US housewives to use the product because it would save them time, not realising that in 1950’s America, toiling away for their husbands was exactly what housewives did. So the product proposition requested a behaviour that contradicted the target’s social identity.

Social identity effects can, therefore, allow us to predict how consumers may react to what we think is a perfectly sensible intervention. For example, the famous Israeli childcare centre’s efforts to charge tardy parents for late pick-ups can be argued as a social identity effect, as ‘dutiful parent’ was replaced with ‘customer’ who could choose now to legitimately buy the product on offer, resulting in more children being picked up late! It also explains why blood donation levels go down when blood donors are paid to give blood – the financial incentive betrays the social identity of being a blood donor, in that the giving aspect of the behaviour is core to the motivation, and so disrupted when money changes hands. Social identity effects can be hugely useful in understanding and predicting not just deviations from normative models of behaviour, but also layering a degree of certainty over many of the more realistic predictions made by descriptive models.

George Akerlof and Rachel Kranton have explored this notion of social identity from an economics perspective in their excellent book, and as a parallel to the blood donor example above, make the point that if you’ve an engaged and committed workforce, offering them a financial incentive to work harder could be the very last thing you want to do. Why? Because the act of receiving payment contradicts the motivation to engage in the behaviour in the first place, resulting in an inferior outcome in terms of performance. Deborah Small and colleagues presented a fascinating paper at BDRM 2014 that arguably supported this point. Women at a breast cancer sponsored run were asked to make short video testimonials to encourage other donors via an online platform. In one condition, they made the testimonials, and in the a second condition they were financially incentivised to make the same testimonials. The latter testimonials received only a fraction of the donations online compared to the former. Whilst a bizarre and intriguing result, we could argue that social identity effects could explain this drop in effectiveness, as testimonial performance was hindered by the behaviour securing a financial reward.

From a practitioner perspective, behavioural economics makes a compelling case for the importance of descriptive models of behaviour (in our consumers or elsewhere). But in that move across to the practitioner field, there’s the danger that the ‘interesting and the novel’ get all the attention – look at what happens when X is done to Y etc. More often than not though, this doesn’t necessarily help us better answer the question when this will happen ‘out there’ in the real world. Whilst it’s not THE answer, the effect of social context on determining how we want to behave, rather than on how we are steered to behave – is an important part of the answer. It allows us to think about the social becoming intertwined with the psychological, that social context is as much the engine of our behaviour as it is the traffic lanes we’re committed to driving in.

Appreciating that we are motivated to increase our ‘social identity utility’ in any given situation presents us with another layer of understanding as to why we do and don’t do stuff, potentially revealing we’re not as fickle and as ‘satisficing’ as we may think. But more than this, when we recognise the influence of social context and our desire to be a part of it, the social identity view opens up effective ways – cost effective ways – of asking for those behaviours.


This blog post is based on an article by the author that appeared in The Harvard Business Review, Feb 2015

Guy Champniss
Guy Champniss is a consumer behaviour consultant and associate professor of strategic marketing at Henley Business School, UK.
Guy Champniss