Highlights from Behavioural Science Summit 2018 at Warwick B-School
For a first year the Warwick Behavioural Insights Team hosted an event bringing together academia, industry, and students at Warwick Business School to discuss the frontiers of behavioural science and its practical application across sectors. Here’s a summary of what happened and my highlights of the day.
The Behavioural Science Summit was organised entirely by the student group Warwick Behavioural Insights Team (WBIT) which was set up in 2017 with the purpose of applying the theory of behavioural economics into action. The selection of speakers included a mix of academics and professionals working across public and private sectors.
Advances in Behavioural Science
The day was opened with a keynote by Nina Mazar who is a Professor of Marketing (Behavioural Science) at the Boston University and co-founder of the centre ‘Behavioral Economics in Action at Rotman’ (BEAR). Previously she held the role Senior Behavioural Scientist of the World Bank’s new Behavioral Insights Initiative (eMBeD) and worked with a range of international organisations. Her engaging talk emphasised the growing influence of behavioural science across sectors — not only nudge units are mushrooming in governments worldwide, but also companies such as Uber and Google are appointing behavioural scientists for their product design.
“It is an exciting time to work as a behavioural scientist” — Nina Mazar
She also showed how applying behavioural insights is shaping as a structured design process and pointed to her Practitioner’s Guide to Nudging. Nina took the audience step-by step through the process by giving various examples from her practice, such as increasing honesty with a pre-commitment field on insurance policy forms, preventing delinquency among credit card holders by testing different automated voice messaging reminders, and reducing corruption in Nigerian hospitals where medicine is often sold under the counter.
Nina Mazar highlighted that behavioural interventions might work in one place but not in another — cultural and social norms play a crucial role in determining the effectiveness. Also people have different mental models for acting in the physical world compared to online, so nudges may not be directly translatable from analogue to digital and more research is needed in this area.
As a closing remark she encouraged to think beyond the effectiveness of a concrete intervention and look into how behavioural design is impacting the wider social welfare.
“It is always important to work in a team with different disciplines that have different types of measures and interventions — other people have good ideas too.”
— Nina Mazar
My key takeaway from this wide-ranging talk is the importance of applying behavioural science theory as a guidance, not as a principle that always works — there are so many factors affecting the effectiveness of interventions that it’s only through experimentation and deep understanding of the context that we can create successful nudges.
Behavioural Science and Financial Regulation
Next up was Paul Adams who started the Behavioural Economics and Data Science Unit at the Financial Conduct Authority (FCA). He talked about the past, present and future of behavioural science in financial regulation. Since he joined FCA his team produced over 15 studies focusing on switching saving accounts between financial providers, overdraft measures, investing fund rates and others. The speaker presented various case studies and highlighted the importance of data science combined with understanding of psychology and consumer behaviour to improve financial wellbeing in society.
“We can get vast amounts of data from people online and analyse it. As a regulator we can then intervene at scale and relatively quickly, so digital is a space where behavioural science can help people.” — Paul Adams
Paul Adams talked about the inertia preventing people from switching to more favourable savings accounts and investment funds. I am personally curious how these customer behaviours will change with the rise of open banking and challenger banks offering greater transparency. I anticipate that the easy open access to customer data (with consent) will level the playing field between customers and providers to enable smarter financial management services as well as easier decision making.
According to Paul Adams, open banking will not have a significant effect in the next 5–10 years, however I already observe interesting behavioural changes on a small scale in my network and will keenly follow developments in the open banking area.
Using Data Science to Quantify the Link Between Beautiful Environments and our Wellbeing
Chanuki Illushka Seresinhe delivered a fascinating presentation of her PhD work at Warwick Business School and Alan Turing Institute. She is using behavioural data science to determine relationships between beautiful environments and wellbeing.
Chanuki took us on the journey of her rigorous research process of using neural networks to analyse 217,000 images from different areas in the UK. Her machine learning algorithms are inspired by the way the brain works and categorise photos as ‘scenic’ or ‘not scenic’ based on data of human ratings.
She differentiated between simply green nature spaces and aesthetic urban spaces to conclude that beautiful environments have a positive effect on happiness and wellbeing. Chanuki is currently extracting urban design characteristics to explore how to design cities that people want to live in and are good for wellbeing. Truly inspiring research with so many applications!
Future Directions in Applied Behavioural Science
After a lunch break the stage took Ivo Vlaev — Professor of Behavioural Science at University of Warwick. He offered a concise walk-through the academic evolution of behavioural science starting with the dual-process theory by Daniel Kahneman and argued based on experimental testing that “for every decision we think hard about, four we make automatically”.
He emphasised that when interventions are based on good theory they are systematically more effective compared to those based on creative ideation.
“There is nothing more practical than a good theory” — Kurt Lewin
Ivo Vlaev showcased practical examples of theoretical principles, i.e. social norms as a principle that humans are strongly influenced by what others do applies as re-framing information from “It is recommended that the average male adult drinks less than 21 units per week and less than 3–4 units daily” to “You drink more units per week than X% of male participants”.
Another example is the ego bias implying that people act in ways that make them feel better about themselves. To increase timely tax payments Prof. Vlaev reframed the messaging from “You don’t need to attend court if you make payment or call us” to “We had treated your lack of payment as an oversight. If you don’t contact us we’ll take it as an active choice.”
A brilliant example of using affect heuristic is the social experiment “Babies of the Borough” — using baby images with round features and big eyes to promote a caring response in human beings and prevent vandalism. To stop anti-social behaviour during the London riots in 2011, local baby faces were painted on the shop shutters in the affected area and showed positive effects.
Lessons Learned from Behavioural Insights Initiatives in the Middle East
Following the theoretical underpinnings of nudging, the discussion moved to implementing it in practice with Fadi Makki — founder of the Qatar Behavioural Insights Unit and the Beirut-based NGO Nudge Lebanon. He introduced behavioural insights at a strategic point in time during preparation for the FIFA World Cup in Qatar. His unit started with small scale nudges such as improving working conditions for workers in stadium constructions and then gradually expanding the scale of experiments to promoting healthy eating, smoking cessation, and reducing motorway speeding.
It was interesting to hear behind-the-scene stories of setting up a nudge unit from scratch. There is no set model — a nudge unit can operate as a governmental agency as well as an NGO depending on the context. It is essential to obtain buy-in from beneficiaries prior to implementation, have close links with academia and appoint an ethical review committee.
How Technology is Shaping our Minds
Next up was Colin Strong — Global Head of Behavioural Science at Ipsos. He delivered a thought-provoking talk on the impact of designing the tools we use which in turn shape our expectations of the world.
We shape the tools, but then the tools shape us and history has shown us this. — Colin Strong
Brand clients often reach out to consultancies with confused challenges and so the real value of behavioural science theory is providing guiding principles to frame the right questions. Without theory we’d be just wandering in abstract space.
Colin Strong also pointed out how using behavioural insights is helping strategists move from focusing on mental states, such as customer needs and expectations, towards designing for mental processes —such as in context experiences.
“We are living in a new context, surrounded by tech and immersed in our screens, we need to understand how we are behaving differently in this online world.” — Colin Strong
After making a strong argument for how behaviourally designed technology is shaping our perception of the world, Mr Strong left the audience with three thought provocations:
- Is the rise of behavioural science a function of tech reshaping our minds?
- Are we in danger of creating over-deterministic environments?
- How should brands & public sector bodies respond?
The Science behind Customer Experience
From brand consulting to customer experience optimisation — Jez Groom from Cowry Consulting explained how he uses behavioural principles to make financial services easy and effective for people. By understanding how customers make decisions, his team considers cognitive biases and heuristics to transform call centre scripts from transactional to relational, improve understanding, make business interactions more user friendly and adapt office environments for effective and enjoyable work.
Making New From Ancient
Sam Tatam — Behavioural Strategy Director at Ogilvy — presented his techniques to influence behavioural change. He pointed out that behavioural science is not at all a new concept but applying it creatively in new contexts is what brings out unseen opportunities. He explained with corresponding examples how Ogilvy’s operating system is based on a foundation of behavioural science coupled with creative thinking and real world iteration.
As a case study the speaker demonstrated the use of visual feedback, negative habit formation and social norms to nudge workers in a food processing factory to wash their hands as required to prevent salmonella infections — something as simple as a stamp on the palm and an instruction to wash hands until it disappears proves to have significantly positive results. That’s the magic of behavioural design — introducing small changes strategically to bring about big impact!
Anthropomorphic Learning
The following speaker took the audience on a ride to the (near) future with a fascinating talk on behavioural AI and anthropomorphic learning. Ganna Pogrebna is a Professor of Behavioural Economics and Data Science at University of Birmingham. In her view, behavioural scientists work in silo isolation and need synergy between disciplines for a more comprehensive tackling of challenges — like a smoothie of different fruits and flavours.
She presented the case for anthropomorphic learning that models human decision making. For example Amazon’s personalised recommendations are based simply on machine learning and don’t seem to represent adequately users’ preferences because people don’t always look for themselves or for what they want. Anthropomorphic learning adds randomness to the algorithm by combining behavioural models with AI to improve predictions of human behaviour in variety of contexts. She tested it with a Tinder-like app presenting shopping deal predictions to participants and vastly improved customer experience with 80% successful matches.
By using behavioural science and combining it with AI and machine learning we can improve the algorithms and improve how we use our technology and our experience on it. — Ganna Pogrebna
In conclusion, Prof. Pogrebna asserted that the future is all about behavioural science and machine learning and encouraged the audience to look into it.
As far as I’m concerned, I’d rather teach AI to think more like a dog than like a human, but that’s my own personal notion on how to make the AI world a better place :)
Panel Discussion on New Frontiers of Behavioural Science
The five previous speakers joined a panel moderated by Prof. Ivo Vlaev. The discussion focused around the ethics of nudging prompted by audience questions. The speakers shared the view that the intention of nudging should not focus entirely on the desired action, but on helping users make informed and considered choices. GDPR in this case helps to level the playing field between the ethically sound academia and the industry practice. As a rule of thumb, we should always use our personal judgement to introduce nudges where the goal worth achieving overweighs potential risks.
Four principles for avoiding misuse of nudges established by Helena Rubinstein were highlighted in this regard:
- Nudges that make it difficult for people to choose otherwise are unethical; consent should not be hidden.
- Nudges that limit consumer choice violate the definition of a nudge.
- Behavioural interventions built on untruths are unacceptable.
- Behaviourally-based interventions should be scrutinised for unintended, as well as intended consequences.
Ethics in behavioural design is an area that deserves more attention from industry practitioners. In academia an ethical review takes around 6 weeks. In the fast-paced design practice that’s often how long an entire project lasts. Designers are usually not formally trained in ethics and there’s general lack of understanding about what ethical considerations entail in design context.
At Common Good we’re working on an Ethics Kit platform alongside Phil Hesketh to bridge this gap by crowdsourcing practical tools that turn ethical principles into practice and guide designers to make considered choices. If you’re interested to learn more and contribute, please sign up here: http://ethicskit.org/
Fireside Chat on Behavioural Science in Finance
Stuart Johnson (HSBC) and Paul Adams (Financial Conduct Authority) sat down for a chat about different aspects of behavioural science in banking. The discussion was ‘kindled’ by Prof. Neil Stewart who introduced a number of studies on credit cards overdrafts, debt management and contactless payments.
As Head of Conduct Risk and Behavioural Economics, Stuart Johnson is responsible for ensuring that the best interest of HSBC clients is considered at every step of the customer journey. The goal is not necessarily making it easy and short, but enhancing customer’s understanding and active choice — i.e. at credit application. He also mentioned an example of informing customers via sms that they are about to enter overdraft just in time for them to make a transfer to the account and avoid fees.
From regulatory point of view, Paul Adams discussed the implications of automatic pension enrolment with a default set too high may result in people actually opting out completely or ending up with more debt. On the flip side auto enrol on a low rate leads to not saving enough for retirement which is also not in people’s best interest. He emphasised the need for rigorous experimentation to determine a sweet spot of positive impact with no backlash.
It would have been interesting to hear the speakers discuss the role of behavioural science in the future of open banking and paint a picture of what behavioural changes are anticipated considering the drastic change in the financial landscape.
The Behavioural Science Summit was a day jam-packed with insights and different perspectives on applying behavioural science in practice. Special thanks to The Warwick Behavioural Insights Team of volunteers for inviting such a great mix of intriguing speakers and for preparing a very useful Nudgelet takeaway with summaries of case studies related to the talks. I love it! I hope the WBIT summit becomes a tradition and looking forward to a 2019 edition!
