Highlights from Behaviourally Speaking, Episode 2

Published in

Behavioural Design

11 min readNov 19, 2018

Last week (14/11/18) we hosted our second episode of the Behaviourally Speaking meetup series.

The meetup’s mission is to bring together behavioural researchers, practitioners, and enthusiasts, to connect and learn more about the field and where things are going.

For more context, you can read this recap of our first meetup, which goes into a bit more detail about the space, and the areas we intend to explore.

Applying behavioural science inside of large organisations

For our most recent meetup, we wanted to explore how behavioural science is currently being applied within organisations, across both the private and public sector.

Outsiders, or practitioners who are just entering into the field, often build their idea of behavioural science on tales from books like Dan Ariely’s Predictably Irrational, or Richard Thaler and Cass Sunstein’s Nudge. Although these are good starting points for understanding the field, they don’t provide a detailed or clear window into the systems, processes, project structures and day-to-day operations of teams working within this space.

The point of this month’s Behaviourally Speaking, therefore, was to really get into the nuts and bolts of how behavioural teams operate, in order to surface the different variations, and treat any misconceptions that might exist.

To help us explore this, we were fortunate enough to have three local leaders speak at the event. These individuals are all pioneering behavioural science initiatives and actively applying behavioural insights within their organisations to solve a variety of different challenges. We mapped out a set of focus areas to guide the speakers’ presentations, as well as the discussions among attendees.

These focus areas included:

Understanding the value proposition (i.e., what needs these behavioural science teams were initially set up to solve) and how this has changed over time
The location and structure of these teams within their organisation
Outputs, outcomes and how individuals and the teams measure success
The different roles and day-to-day operations of individuals within these teams
Required competencies for being effective within these teams (general and technical)

Ammaarah Martinus

Director of Policy, Research and Analysis, Department of the Premier, Western Cape Government

Ammaarah began her talk by showing a well-known map of all the behavioural teams working in public policy around the world. The usual story that is told in conjunction with this illustration, is of the immense growth in applied behavioural science around the globe. However, Ammaarah then invited the audience to consider an alternative interpretation. Looking at it differently, the map also exposes the limited amount of formal activity in Africa and South America, when compared to the rest of the world. Besides a collection of consultancies, research labs, like Busara and government units, like the WCG’s and the EiP in Colombia, there isn’t much activity happening in the Global South. Of course, what this also points to is the immense potential for rapid growth over the next few years; especially as teams in countries leading the way start getting recognised. Ammaarah wants her team to be at the center of this growth phase.

With the mission explained, Ammaarah went on to talking about the WCG’s history working in the BI (Behavioural Insights) space, and the immense role that partnerships have played. From the very outset, working with UCT’s RUBEN and ideas42 in 2012, to their more recent work with the World Bank and OECD, working closely with their partners has always been at the core of what they do. The benefits of this were evident at the recent OECD-WCG Behavioural Insights conference held in Cape Town, during September this year. Collaborations between large global research teams and local practitioners seems to be a core characteristic of the work being done in this space.

From the team’s mission and partnership structures at a high level, Ammaarah then delved into the day-to-day details of the work her team is doing and where BI plays a role. She neatly organised these day-to-day activities into three general themes:

Policy development — moving behavioural insights further upstream in the policy cycle, with the aim to have BI included in policy creation as opposed to just policy iterations and operationalisation
Planning and implementation — designing interventions using existing behavioural evidence, trial design, implementation and evaluation
Stakeholder management — looking at how behavioural insights can be baked into existing government processes, communications, meetings and workshops

The roles and structures required to execute these activities were described as flexible and adaptable, but again three general roles emerged, as necessary for running a BI team within a local government:

The Director — responsibilities include: partnerships, fundraising, strategy development, stakeholder management, and intervention design input
The Project Manager — responsibilities include: stakeholder management, field management, and intervention design input
The Project Coordinator — intervention design input, project execution, and stakeholder management

More informally, Ammaarah suggested seven types of roles that need to be covered in order for BI work to be successful in the public sector:
(The orchestrator, the advocate, the steward of place, the entrepreneur, the catalyst, the buffer and the sensemaker)

In summary, a flexible team that’s made up of a variety of competencies needs to be able to work closely with different kinds of implementation partners (universities, consultancies, other departments) in order to influence policy decision making, design interventions, run trials, manage data and evaluate the impact of interventions. Once this complexity is appreciated, the need for a strong collaborative mindset becomes distinctly apparent.

To end off her talk, Ammaarah shared some take-home guidelines for embedding BI in the public sector:

Guidelines:

Use BI to supplement and support existing programmes
Consider the policy layer before starting the intervention
Use partnerships to crowdsource input
Put a political and administrative champion in place
Make sure you have a sufficient amount of time to run BI trials
Make sure to have access to accurate administrative data
Invest time in navigating the procurement and IT processes
Make sure scale-up plans are set and committed to before a trial kicks off

Kevin Joseph

Behavioural Economist in the Behavioural Economics Team (BET) at Old Mutual

What is ‘the need’ for behavioural science teams within organisations and what value are they adding? This was where Kevin kicked off an interesting and detailed deep dive into the inner workings of private sectors BI teams working within large organisations.

Kevin was quick to acknowledge the role mainstream recognition of the importance of behavioural science driven by things such as Kahneman and Thaler’s Nobel prize wins, has played in building an appetite for applied behavioural science. However, besides being a foot in the door, Kevin noted that these external forces haven’t done much work when it comes to justifying the team’s day-to-day value within the organisation. Inside Kevin’s organisation, numbers do the talking. This is typically thought about in terms of ROI (Return On Investment) — the quantifiable value generated by a BI project, in relation to the resources invested in making it happen. Positive ROI on a project improves the reputation of the team, opens up access to more resources, more projects and more influence within the organisation, leading to more opportunities to further improve the team’s returns.

In a similar way to Ammaarah speaking about BI moving further upstream in the policy cycle, a goal of Kevin’s team is to move away from BI just serving an optimisation function, and towards a role where they are injecting their insights and ideas directly into the conceptualisation of new products, services, programmes, and communication campaigns. The prospect of BI being more than just an optimisation tool, is something practitioners often talk about as one of the major goals for the field going forward.

Kevin then went into a bit more detail about his team’s objectives, their day-to-day activities and his view on the competencies required to be effective within a BI team.

He nicely summarised his team’s objectives as the following:

To help customers make better financial decisions and improve their overall experience (measures of success: investment contribution rates, retention, re-investment rates, and customer satisfaction scores like NPS (Net Promoter Score)
Improving the wellbeing and satisfaction of employees within the organisation (measures of success: retention, wellbeing and employee satisfaction and productivity via KPI’s (Key Performance Indicators)

Unpacking his team’s day-to-day activities, Kevin emphasised the importance of project, time and stakeholder management; as each individual on his team is almost always working on more than one project at a time, with tight deadlines, and a variety of different stakeholders from different teams within the organisation. Typically these projects follow the conventional behavioural design process — defining objectives, understanding the decision environment, designing and building project deliverables and running pilots to evaluate the impact of these deliverables.

When it came to the competencies that make individuals effective in a BI team like his, Kevin emphasised the value of domain-specific knowledge, typically acquired through academic training; technical skills, like being able to easily navigate survey/experimentation platforms; and more general softer skills like, stakeholder management, collaboration and the ability to effectively communicate complicated ideas within the organisation.

Jane Ball

General Manager: Strategic Advisory Unit at Medscheme

Jane has been actively applying behavioural science to solve healthcare challenges with a core focus on chronic disease. She began her talk by neatly summarising her problem space and the very clear value proposition for a more behaviourally-informed approach.

Jane built on this by outlining three strategies that have been proven to improve chronic disease outcomes:

Adherence to “best practice’ treatment
Coordination of care
Self-management / lifestyle change

There is a clear relationship between these positive health outcomes and behaviour change. Nonetheless, from a broader perspective, Jane was quick to emphasise the importance of understanding whether a particular challenge is systemic or behavioural before applying behavioural science.We are often tempted to jump to the behavioural component, when in fact, as Dilip Soman often says, we should make sure ‘the pipes are clean’, before considering approaches that aim directly changing client behaviour.

Once a challenge has been identified as behavioural, Jane suggests that the next step is to look at is whether the behavioural problem is one of intentions or actions. As is well known in the behavioural sciences, the sweet spot for applying behavioural insights is in the pool of challenges where the aim is to close the gap between people’s intentions and their actions (Jane calls these individuals inclined abstainers).

Jane drew the audience’s attention to the fact that less than 50% of intended health behaviours are realised. This problem,at its core, has a lot to do with behavioural hurdles, such as competing priorities, immediate costs of action, procrastination, forgetfulness, failure to recognise opportunities to act, as well as control and self-efficacy. What is immediately clear from these hurdles is how little they have to do with changing intentions, attitudes or preferences. Accordingly, this exposes the limitations of educational, informational, or purely persuasive approaches to changing behaviour, as people clearly want to act, but they just don’t get around to it.

Her point was directed at the limitations of information in changing health behaviour specifically, but the insight can be generalised to other domains such as financial behaviour, organisational behaviour, environmental behaviour and perhaps ironically, even educational behaviour. It’s a point that is really at the heart of what our field is all about.

Once Jane had positioned behavioural science and the role it can play in closing health-related intention-action gaps, she went on to share some of the practical resources for crafting behavioural solutions. These included BIT’s EAST, MINDSPACE and UCL’s Behaviour Change Wheel.

Jane was quick to emphasise that these resources were just the starting point. That they could provide some direction, but aren’t silver bullets for achieving the desired health outcomes. In order to know for certain whether an intervention has been effective, you have to run randomised controlled trials. As anyone who has run field experiments will know — things seem easier in theory than they are in practice. Dealing with humans in uncontrolled environments, spillover effects and getting business to prioritise validity were some of the challenges Jane mentioned. The challenges seem to come up fairly often within the context of business experimentation.

It is also important to be reminded, that these challenges and costs dwarf in comparison to the consequences of trying to correct an ineffective intervention that has been scaled before testing.. The Scared Straight programmes in the US are always a good example of failed interventions, if you require some ammunition to make this point more salient. Otherwise, Frank Lloyd Wright’s famous quote gets the point across too — “You can use an eraser on the drafting table or a sledgehammer on the construction site.”

Jane spent the last few minutes of her talk speaking about her team’s future. She neatly outlined three exciting areas that they are currently exploring:

Combining behavioural insights with predictive modeling to improve the tailoring and targeting of their interventions (personalised nudging)
Investing in technologies that allow interventions to be delivered at the most opportune times
Focusing on more ways that choice architecture might be used to facilitate better client decision making

Looking forward

Besides the valuable lessons shared by our three speakers, what was also great to see at the event was the number of questions and commentary from the audience. The discussion points ranged from strategies for applying behavioural insights ethically to the long term effects of extrinsic incentives. These questions provided us with some interesting ideas for themes going into 2019, where we intend to build on the momentum we have generated with this year’s meetups and host more events. I look forward to seeing you there!