Nudge units aren’t wands

The trend for governments and corporates to establish their own “nudge units” continues apace. Our experience from these developments highlights the risk of the behavioural sciences being viewed as a wand in the hand of Harry Potter: a magical instrument for getting muggles to do whatever we want.

The recent publication of McKinsey Quarterly’s Lessons from the front line of corporate nudging in January 2019 brings attention once more to the fact that many corporates are building their own in-house teams of behavioural scientists. This follows in the tradition of other publications, such as the Harvard Business Review’s Building Behavioral Science Capability in Your Company in 2017 as well as similar articles since 2014 which have tracked the rise of the Chief Behavioral Officer. In terms of public sector developments, the OECD’s map on Behavioral Insights and Public Policy traces the expansion of nudge units across the globe.

Consistent throughout are themes of the importance of the behavioural sciences and the need to internalise the capability for strategic competitive advantage.

Our view is that this trend is overwhelmingly positive. It recognises the contribution that this eclectic discipline has to offer to large scale social issues, as well as commercial challenges. Increasingly it is being clawed back from academic institutions to where it arguably belongs: in practice. It signifies, what I refer to as, the “democratisation of the behavioural sciences”.

Importantly, an investment in building one’s own behavioural capability, also – in principal – marks an implicit commitment to more rigorous, fact-based management through a test-and-learn discipline. This point is worth pausing on and rephrasing. A nudge unit is a capability and an approach. But it is also a positive mindset of the value of learning and managed failure. Where this is matched with an organisation’s complementary culture, one can expect … well … not magic, but fireworks perhaps.

In contrast, misalignment is fertile conditions for a Forbidden Forest, a mysterious and dangerous place. This I see to be the negative side of the trend towards building nudge units, where the unit and, by implication, behavioural science, is often seen as some sort of magical wand. Wave the magic wand at a problem and, Evanesco!, problem gone.

The difficulties with this view of the behavioural sciences are several. For one, it creates vastly inflated ideas of what can be achieved through the behavioural sciences. For another, it suggests that behavioural design always achieves the intended outcome, as a result of which testing is unnecessary. Thirdly, it makes the work dangerously formulaic, with the solution space assumed to be uncritically replicable, copy ‘n paste, across different contexts and markets.

The fourth difficulty with the notion that behavioural science is a magic wand is that it is inclined to gravitate more towards packaging problems, and less to the package itself. (Soman’s language for this is last mile, versus first mile problems). This, for me, is especially pernicious for the enterprise as it falls short in delivering on some of the more meaningful, sustained behavioural change that lies beyond simple prods, pokes and nudges.

My advice to internal nudge units thus comes in the form of a warning. Being viewed as a wizard with a magic wand is flattering in the short-run. In the long-run, however, your route out of the woods will be the measurable results that you have achieved through control trials, testing and learning.