The Behaviour Business. Richard Chataway

      From here, the BIT has grown to have a presence in five countries globally, and the model has been adopted in a widespread fashion elsewhere, with the creation of numerous ‘nudge units’. This growth has been driven by a scientific application of the findings of behavioural science.

In government, where accountability is paramount, this evidence-based approached has proven revolutionary. As Halpern puts it: “This is just a glimpse of an exciting and important future: where policy and practice is based on hard evidence, not just instinct or history, and where public money can go further and outcomes continually improved.”25

      Guided by successes such as the reduction in smoking prevalence, governments have moved from a narrow application of the insights from behavioural science in (social) marketing to a much broader, scientific, outcome-based application. There is much that business can learn from this approach.

Nudging through technology: My QuitBuddy

      In 2012, I moved to Australia to take up a role as strategy director at the media agency (UM) for the Australian Federal Government.

      As in the UK, smoking was the single biggest preventable cause of death in Australia. And despite a long heritage of effective behaviour change campaigns giving Australia one of the lowest smoking rates in the developed world, smokers continued to smoke despite being aware of the risks. It was clear, as in the UK, that focusing on the desired behavioural outcome (getting people to quit and stay quit) would be more effective at reducing smoking rates than simply giving rational reasons to give up – giving them the how, rather than telling them why. It would also be considerably more efficient (i.e. cheaper) than an expensive advertising campaign – an example of what Thaler calls “making it easy”, his three-word summary of Nudge.

      When I joined, my new colleagues at UM had talked to our government clients about using then-new mobile app technology to help people quit. Bringing insights into effective ways of nudging behaviour – such as the importance of social proof (i.e. seeing that others had successfully quit using the app) and saliency (i.e. providing bespoke information to each user) – we built an app with development partners The Project Factory called My QuitBuddy.

The first version was built in a mere eight weeks and was very much based on an MVP (Minimum Viable Product) approach,26 with fairly limited functionality. It included motivational messages of support, a game that smokers could use to distract themselves when experiencing cravings, the ability to record motivational messages from loved ones and provided up-to-date data each time a smoker opened the app on how much money they had saved, toxic tar they had avoided and so on.

      After launch, My QuitBuddy quickly achieved the number-one ranking in the Health and Fitness category on the iOS app store, with over 100,000 downloads in the first year. Seven years on, it is still going strong. It has been downloaded over half a million times, and the quitting success rate of users is eight times higher than smokers without support. The app has been white-labelled for use by a number of other governments and is still a key part of Australian government stop-smoking campaigns. It is probably the most effective stop-smoking intervention employed to date by the Federal Government.

      Why has it proven so successful? The initial insight behind developing an app was that most quit attempts fail because cravings can hit at any time, so timely support needs to be within arm’s reach at any moment. For most of us, the only thing within arm’s reach 24 hours a day is our mobile phone. The technological solution was therefore built around an insight into the desired behaviour – and not created for its own sake.

      But, more importantly, we were able to create a more effective, addictive and usable app because we had data on what parts of the app people were using. It has been continually updated over the last seven years based on data on actual behaviour.

For example, we found app users were screen-grabbing the homepage (which showed how much money they had saved, how long they had been smoke-free, etc.) and posting it to social media accounts. In the spirit of making it easy, we updated the app to allow users to directly link it to their Facebook and Twitter accounts, so that with one tap they could post an update. This leveraged two behavioural biases: commitment bias (making a public commitment makes us more likely to stick to a behaviour, in this case telling all your friends you have quit); and social proof (showing how popular the app was would encourage others to download and use it).27

      The app was continually optimised based on actual user data and behavioural science best practice, a distinct advantage over a traditional, one-hit advertising campaign.

      I think this example of how behavioural science has been used to address an important societal problem can tell us a lot about how to solve business problems, as well as save lives, and has been hugely informative in my own work. It tells us about the importance of focusing on (behavioural) outcomes, like helping people quit rather than simply telling them why they should. Similarly, it shows us how behavioural science can lead to a better evidence base and enable more creative solutions (such as the fatty cigarette campaign), as well as how technological solutions work best when grounded in a behavioural insight, rather than simply a desire for novelty or innovation. We shall explore this more in part two.

      In the next chapter, we will see how this kind of scientific approach – based on testing, learning and optimising – is fundamental to building a successful behavioural business.

      19 NHS England, Statistics on Smoking – England, 2018.

      20 www.thirdsector.co.uk/change-makers-british-heart-foundation/communications/article/1192942

      21 If you are curious, Waters told me that the ‘fat’ in the cigarette was made using a combination of hummus and wallpaper paste. Which must have smelt disgusting.

      22 In part five we shall see that the motivational messaging subsequently centred on the consequences of smoking on smokers’ families.

      23 However, this does not necessarily increase the number of people who receive organ donations. The issue is a complex one, as it is also heavily determined by the processes employed by hospitals when a donor dies – as they need to then obtain consent by the next of kin, and have the right staff and logistics in place to ensure the organ can be received and transported in time to be used by the donor. The single biggest improvement in live organ donation success rates in the UK was actually achieved by having a dedicated donor nurse active in each NHS hospital. Obviously having more people on the register is helpful to increase the pool of donors, but the debate is whether a mandated opt-in (so that everyone has to state their preference, and let their next of kin know their intention) is preferable to an opt-out system, which leaves potential doubt about their wishes in the event of their death, and so leads to more next-of-kin refusals. Scotland and Wales have moved to an opt-out system in recent years, and at time of writing England has stated an intention to do the same.

      24 The now defunct COI (Central Office of Information) was established shortly after the Second World War, as the successor to the Ministry of Information (the propaganda department), initially to inform the nation about the newly created NHS. It is perhaps best known for producing much-loved public information films such as ‘Charlie Says’ and ‘The Spirit of Dark Water’, as well as for well-respected evidence-based best practice in behaviour change communications. Despite its integral role in the creation of the BIT, it was closed by Cameron’s government in 2012 as part of austerity measures. Following a public vote of no confidence in its successor by the Institute of Practitioners in Advertising (due to its procurement processes), a new Government Communications Service was subsequently established – effectively a COI 2.0. Meanwhile, nearly all COI staff had either left (like