Collections might just be the most rewarding field to work in for a product designer or behavioral economist. Elizabeth Immer explains why and reports on how the Ergonomen-guided overhaul of Swisscom’s collection policy led to savings of more than CHF 8 million in the first 2 years.
When I tell people that working in collections is the “best topic ever,” I get some confused reactions, to say the least. And I get it. At first glance, collections appears to be a little bit yucky. But in my experience, working in this branch is anything but. It’s exciting and rewarding! Exactly because so many parties stand to win when collections “works better.”
Using the tools and learnings behavioral economics offers us in order to make a difference, to help people “do better,” gets me out of bed every morning. Thanks to a collaboration with Swisscom’s billing team since 2017, I’ve had the opportunity to do just that.
Before getting into the story, I’ll cut to the chase. Swisscom saved more than CHF 8 million (in 2 years) after we helped them revamp their collections policy. Their customers in collections are markedly happier than they used to be. And the number of customers whose contracts are cancelled due to unpaid bills has sunk by two-thirds.
I’m writing here to share the story behind this happy result – how did we get there? We tend to think of people who don’t pay their bills as bad, delinquent folks – unattractive “others”. You aren’t one of them, are you? Me neither, I’m perfect. Just kidding! Most of us are imperfect in this sense.
In Switzerland (and not just here), quite a lot of people fail to pay at least some of their bills on time. This trend is getting worse, not better. In 2017, after many years of systematic processual changes, Swisscom’s billing department had the feeling they could be more successful at getting their customers to pay their bills less late. They asked us to help get their customers to behave better, while keeping them happy.
I was excited by this question because I love challenges exactly like this: How can a product be tweaked to help people change their behavior substantially without tricking, frustrating or otherwise bothering them?
Of course, it really doesn’t make sense to not pay bills. When asked in a cold state, everyone acknowledges this. People know that bill paying isn’t optional. Otherwise they end up in debt, with destroyed credit scores. A “perfect person,” sometimes called Homo economicus, will recognize this and will always pay her bills. The problem is that humans are not perfect in this way. (Behavioral economists see this over and over again: We people systematically fail to do what we should do, and often what we want to do.)
As nearly all companies do, Swisscom penalized customers whose bills were overdue, including by imposing fines. It got expensive to be incompliant. But despite these penalty fees, people weren’t paying. The penalty fees acted as just that – penalty, punishment. They made customers mad without getting them to change their behavior. But while we realized that penalties weren’t working, it wasn’t clear what else could be done to systematically change behavior.
Our first goal was to understand this “why.” Why are people doing what they are doing? We needed to collect evidence to get to the bottom of this. Our second goal was to design a new policy that worked better. Sounds easy, right? I’ll admit – I had some sleepless nights during the design phase. But the tools that my background in behavioral economics have equipped me with served us well, enabling us to achieve these goals, with considerable impact.
How did we tackle these challenges? First, we asked “why aren’t people paying their bills?” In this user research phase, we ran in-depth and often quite emotional interviews with customers who had experience in Swisscom’s collections process. We visited call centers and spoke with call center agents, who speak with customers all day every day. As in every project, we also dug into scientific literature, especially from behavioral economics.
Of course, the topic was new to us and the existing processes complex. We couldn’t know everything! Not only for this reason, the billing team-members (our clients) were our most important enablers. In order to work effectively with them, it was essential that we communicate our findings (also the complicated scientific concepts) in a way that all the stakeholders (who had no background in behavioral economics, human centered design or related fields) could understand and work with us easily. The simple behavioral model we defined based on all that we learned from the existing scientific research and the context-specific user research phase was a powerful tool in this regard.
In looking back, I give the focused research phase credit for much of the program’s success. After the research phase, it was “relatively easy” to draft a new process. (“Relatively easy” = those sleepless nights I referred to above.)
In the design phase, we overhauled much of the policy. The policy? Yes – essentially the whole experience a customer has when she doesn’t pay her bill on time: When will she be contacted by Swisscom? On what medium? How will Swisscom incentivize her to pay (with carrots or sticks or both)? What should the emails look like? What words (and numbers) should be used? All with the goal of having the greatest possible positive impact on behavior.
In the design, we considered not only monetary costs and benefits, but also psychological and social costs and benefits. After all, we behavioral economists know that money matters, but psychological and social factors do too. Here, I mean aspects like expected ease/difficulty of doing something and social status or embarrassment can also be important drivers behind behavior. Of course, we also considered the powerful heuristics at play in our minds. Heuristics, AKA shortcuts that our brains take that make it possible for us to process and respond to everything we are faced with each day, are responsible for biasing our perception of incentives, sometimes leading to poor behavior. Behavioral economists focus on understanding what costs, benefits and heuristics systematically lead us to behave the way we do (in what ways we are systematically “imperfect”). We took all these aspects into account when designing the new policy.
We tested the impact of the new policy in simulations in our lab, using high-fidelity prototypes. And of course, the extended project team including Ergonomen and Swisscom billing team members then refined the products based on what we learned in the user tests. After we were confident with the new version, it was rolled out to a few thousand customers in a pilot phase run by Swisscom.
In other words, our process has been hypothesis led, empirically informed, collaborative and iterative. Those are a lot of adjectives, but empty words? Anything but. This approach has helped us all to not only work smart but also to be very brave in our designing out of the box thinking. Oh, and to have fun 😊.
I find it rewarding that our project hasn’t just led to an improvement in a single process, but rather, that there have been positive ripple effects in our customers’ organization. Since our collaboration began, my colleagues at Swisscom have become more and more engaged in user centered design and behavioral sciences themselves, with one person even enrolling in a continuing education course in this field!
FRESH, the name of the overhaul program reflects (or maybe caused, hard to say) this enthusiasm. We named this program Fresh as a nudge for everyone working on the project. The name was inspired by work by Katy Milkman around the “Fresh Start Effect.” Katy’s experiments on behavior change and goal achievement demonstrate that it can be productive to start something new on the first day of a season, month or week. We weren’t willing to limit the “Fresh Start” mindset to one day, though. We chose the name to help us every day to remember our goals of working with a clean slate, of being willing to challenge assumptions that had been held onto for years and to be brave while reconsidering and redesigning this important product. I guess it’s safe to say that the nudge worked as hoped.
Photo credit: Louis Rafael Rosenthal