The disfluency of change

Sometimes we need friction to disrupt how we process information and to make change happen

Changing behaviour is uncomfortable – we often have to push ourselves out of old routines, force ourselves to consider alternatives. This is understandable given the way our behaviours reflect the effort we have previously put in to look at options, work things out, to have arrived at a behaviour which suits us.  We can then process our environment in a ‘fluent’ manner, where we are familiar with the options and how we respond to them.  Change means uprooting and challenging ourselves:  it means creating ‘disfluency’. 

Of course, small changes tend to ‘go with the grain’, not requiring much, if any effort.  Indeed, the point of nudges is to change behaviour intuitively without us feeling the pain of change.  But increasingly we are being asked how to launch services, offer products, create public health or finance programmes that are not incremental but result in a step-change in behaviour. 

It is the discomfort or disfluency that results from changing behaviour which can actually be our guide for how to navigate it.  If people feel ‘disfluent’ then they are re-examining their options, reflective and curious of the alternatives, they no longer feel a sense of ‘rightness’ about the way they currently do things.  It is just this state that we need if we are to disrupt current behaviours and encourage change.

The value of disfluency

Sometimes we cannot always run an experiment to assess the impact of an intervention on behaviour – perhaps it is simply not viable such as trialling a new cereal brand.  We could seek out a trail supermarket or even country but the logistics of setting up a field trial are not only onerous but there might be budgetary or other logistical limitations

So the typical route is to then undertake some kind of test, typically online, where some are shown the ‘intervention’, in this case let’s say advertising for a new breakfast cereal, while others do not see anything (or other variants).  This is an area in which there is a long and well established set of protocols in market research with a range of outcome measures that can be examined to assess any differences between the experimental and control groups. Testing for incremental change is well understood and there are very effective research protocols for this.

The challenge comes when testing stimulus which represent a step-change to normal behaviour.  For example, perhaps the cereal has a personalised element to suit your specific needs.  Or there is a refill option available which involves exchanging the packaging for a new pack. 

How can behavioural science be used to assess the impact of these, fundamentally different, propositions which require a very different set of behaviours?  One approach we have been developing is to assess the degree to which the proposed intervention is disrupting the way people are thinking and slowing them down.  At first glance this seems at odds with received wisdom – surely we should be making it easy? This is an approach frequently cited, including via the well known EAST framework.

But do we?  Remember that if we are looking to make incremental changes then perhaps we want them to be easy and for it to be aligned with the way in which people typically process information.  This is the premise behind nudge – go with the grain of people’s processing and allow the choice architecture to gently move behaviour to one side, perhaps even without the participant being aware.

But when you have something that is more disruptive, then this choice is not available.  In fact, we can question the assumption that we want people to process information in a fairly intuitive way.  We want to instil some disfluency if we want people to change.  People need to engage and rethink their current behaviour and assess this new option.

This is also the case for a market leader versus a brand challenger.  A market leader will likely want people to continue to process information in a fluent, intuitive way: after all, the way people are behaving is suiting the brand leader well and incremental changes can then be made with new variants and brand extensions.  A brand challenger, on the other hand, wants to disrupt processing as the way it is currently done does not benefit them.  This disruption, or disfluency, encourages people to slow down and reflect, reconsidering the choices they are making.

Behavioural science can augment current tools in these circumstances, offering more understanding about the mechanisms underpinning change and helping to evaluate the way in which the proposition is creating a readiness to change.

Disfluency as an impact measure

With this in mind, there are implications for how we can go about assessing the impact of interventions.  If we are interested in making incremental changes, relying on reinforcing or making slight adjustments to the way in which people are processing information and making decisions then we want to see fluency, an ease of processing.  If we are seeking more substantial change then we want to find disfluency, more deliberative processing. 

At this point it is reasonable to question what constitutes incremental from more substantial change. And as might be expected, there are no clear answers here, it depends on context.  What is incremental for one group or category may be business as usual for another.  Nevertheless, hypotheses can be made.

Next, we need to consider how to measure the fluency-disfluency continuum.   Measuring response times is one option:  in a twist to the usual assumption that fast is always best, we can consider that a slow response may at times be what we are looking for when considering a disruptive new product, as it reflects the way our processing has become disfluent.

But given it is not always feasible to measure response times, we use a series of questions relating to fluency / disfluency:   these include curiosity about alternative solutions, perceived ‘rightness’ of the current solutions and ability to explain why current solutions are preferred.  These measures offer a way of assessing the degree of fluency – disfluency so we can identify how the proposition (or intervention) is creating a readiness for change.  Alongside this we can examine the barriers to actually changing using our MAPPS framework, locating the different types of barriers that need to be mitigated in order for change to happen (which then link through to intervention strategies).

Conclusions

We often want things to be frictionless and easy for consumers:  indeed, often our focus is on reducing friction to enable outcomes.  While many benefits have been seen from facilitating fluency, we argue for a more nuanced perspective; one that recognises the utility of creating a frictionless set of options where a smooth, easy and convenient experience is optimal, but also creates friction when deeper processing necessary for changing behaviour is warranted.  If we are seeking incremental change then ‘nudges’ can help deliver this in a fluent manner.  But if, as we are increasingly seeing, the demand is for more disruptive change then we need to find ways to introduce more friction and disfluency to help people to examine options and properly consider change. And to evaluate the way in which interventions as well as new products and services deliver on this, we ourselves also need to think carefully about how to augment the measurement tools we use.