Born on Christmas Day 1935, Donald Norman has dedicated his professional life to design, usability engineering and cognitive science. He operates in a world where people, work and technology connect, or more often, fail to do so – a world known as cognitive systems engineering.

Like few others, Norman has shaped how consumers interact with technology. One of the first markers he laid down came in a 1981 article published in Datamation - “The truth about Unix: The user interface is horrid,” turned him into a computer industry celebrity. From there, Norman would introduce the world to user-centred design, a term he coined in his 1986 book User Centered System Design: New Perspectives on Human-computer Interaction. “People are so adaptable that they are capable of shouldering the entire burden of accommodation to an artifact,” he wrote, “but skillful designers make large parts of this burden vanish by adapting the artifact to the users.”

It’s easy to forget how radical an idea this was at the time – that computer interfaces could be designed with users in mind. It would be another seven years before Norman was hired away from academia by Apple Computer as a user experience architect. His job title is said to be the first ever to feature that now common phrase.

Traditional vs. behavioural design

The principal difference between traditional and behavioural design is that while the former solves functional problems with aesthetic and usability improvements, the latter seeks to affect human behaviour. It does so by studying users, testing options and honing the experience to produce outcomes that are in the user’s best interest.

The Canadian pension industry’s reliance on behavioural design – and the related field of behavioural finance, which has taught us not to rely on the traditional view that people are good at making rational decisions – has enabled measurable improvements in plan member experience and retirement outcomes in recent years. We owe pioneers like Norman a real debt, one that probably deserves greater appreciation than it’s received.

More recently, the application of machine learning and artificial intelligence (AI) has brought about a step change in member experience best practices. Algorithms equipped with logic, reasoning and conversational abilities present us opportunities that were unfathomable just a few years ago. At the same time, many plan administrators have invested in upgrading legacy infrastructure, creating data strategies and building teams to take advantage of AI.

It is early days yet, but there is reason for optimism about the future of retirement plan member experience. In four important ways, insights from the fields of psychology, behavioural economics and design thinking can now supercharge plan member engagement with the help of AI.

1. Generating interest

Typically, a member’s initial point of contact with the plan is a kind of registration process. It’s a little like being handed a new patient form at the doctor’s office. As first impression’s go, we can do better.

The world is full of distractions, many of them digital. TikTok is never more than a click away and we are competing with that for members’ attention. All plan member experiences – including the onboarding process – should be captivating and foster curiosity. It is after all a genuinely valuable thing they’re being presented with. Applying a behavioural design approach to member experience entails observing how members navigate any given process as well as analyzing quantitative and qualitative data on the steps they take (and elect not to take).

This more detailed, data-driven approach to experience design presents us opportunities to make adjustments, and then test and learn how effective they are in generating interest and prompting member behaviour that’s in their best interest.

2. Decision support

The burden of decision making – particularly in defined contribution plans – is on members, many of whom don’t understand even basic financial concepts. They’re required to make choices they’re uncomfortable with, which in turn produce long-term results they can’t foresee.

Generative AI combined with optimization algorithms can simplify the member experience, based on their individual level of financial literacy and where they’re at in their retirement planning journey. 

There can be different conversational AI models for accumulation and decumulation stages, and for people with different levels of financial knowledge and interest. This can apply to a variety of retirement decisions, such as buying back service, commuted value calculation, asset transfer, etc.

3. Eliciting positive emotions

Behavioural design is often associated with usability. In the context of pension plan experiences, the goal is to provide members an understanding of how the plan works, and if they’re required to make choices, ensure they have the information they need to make them on an informed basis.

But pension plans have a dual purpose – they produce retirement outcomes and, ideally, serve as a talent attraction and retention tool for employers. To achieve both, interactions with the plan should be more than simply functional. They should trigger an affirmational response on the part of the plan member, the kind that leads them to think more positively about their employer.

This is a good example of how generative AI is making behavioural design so much more effective. The technology genuinely engages members, both personally and responsively. It’s a powerful improvement relative to what they’re accustomed to.

4. Behavior initiation and action guidance

Chatbots powered by generative AI are adept at engaging plan members in productive conversation. This captures valuable insights into the plan member perspective. It fleshes out their expectations, worries and fears. 

The technology can provide rapid feedback on behavioral insights, which can be used to fine-tune conversational AI models for more engaging member communication. This creates an opportunity for administrators to ensure that members truly understand and appreciate the value of retirement benefits. That helps remove barriers to action and initiates positive action.

Experts in behavioural experience point to a range of behaviour change principles. The list includes personalization, education, feedback, relevance and empowerment.

On each of those points, AI-powered behavioural design makes it possible to develop pension plan experiences that support a superior level of member engagement and stronger habit formation. 

Members are presented with their own data and relevant educational information in a manner the administrator knows will be most effective. They receive real-time feedback that supports sound decision-making and leaves the plan member with a greater sense of enablement.

Real tendencies, not best intentions

Behavioural design represents an inherently human-centric approach. Rather than focusing on how people should behave, it emphasizes their actual tendencies. That’s important because so many of our decisions are driven by subconscious impulses and deeply ingrained habits.

In 1979, Norman was invited along with a group of academics to investigate the Three Mile Island nuclear accident near Harrisburg, Pennsylvania. The facility’s Unit 2 reactor suffered a partial meltdown after a series of mechanical failures including a pilot-operated relief valve that got stuck open. In the language of nuclear energy, it was a loss-of-coolant accident. Radioactive gases and iodine were released into the local community. 

What Norman and his colleagues learned was that operators were poorly prepared for the scenario, and that numerous design flaws exacerbated the crisis. It was a kind of perfect storm that failed to account for the very real tendencies of human beings in an imperfect environment. It’s a dynamic we continue to see in a wide range of circumstances, even as we’ve made steady progress on behavioural design.

In many respects, 1979 feels like ancient history. Even those of us who saw it up close recognize that we’ve come a long way on multiple fronts. For pension plan sponsors, advances in behavioural design, and more recently generative AI, provide reason for real optimism about retirement outcomes and the effectiveness of these plans as talent attraction and retention tools.

Leyla Imanirad, Co-founder & CEO, Pensionbar

Leyla is the co-founder & CEO of Pensionbar, a fintech company based in Toronto. The company’s mission is to enhance member experience through combining technology, data science, behavioural finance, and user experience. As a technical advisor to different companies and projects, Leyla raises awareness about the impact of AI on different industries. She has a master’s degree in computer engineering from University of Toronto.

Saman Khodai, Co-founder & Director, Pensionbar

Saman is the co-founder and director of Pensionbar and the founder of Pension Transitions, a pension consulting company. He believes in the power of technology to simulate & analyze corporate DB pension plans under different scenarios. As a Certified Professional Retirement Coach, he consults individuals and corporations on retirement wellness and readiness. He has a PhD in pension finance from university of Graz, Austria.