Platforms and user-centric design

“Every time we make a choice, someone has been there before us. We’ve had a hidden partner who’s made a bunch of decisions.” These are the practical terms that Professor Eric J. Johnson uses to describe choice architecture and its powerful influence on our digital experiences. Eric is Director of the Center for Decision Sciences at Columbia Business School and an expert in choice and decision-making, and as he points out, there is always a framework behind the choices that we’re offered.

Choice architecture is necessary; even its seeming absence is influential. Eric uses Google Flights, which displays environmental impact metrics in addition to basic information like price, as an example. “They’ve made a decision […] to include the carbon emissions for the flights. Now, that’s going to have an influence. There’s no neutral – you either have that [information] or you don’t.”

 

Gathering insights into consumer behaviour

According to Elisabeth Costa, Managing Director of The Behavioural Insights Team UK, “[platforms and online environments] are highly curated and deliberately designed,” and they offer a unique opportunity to observe and measure users’ actions. Elisabeth explains: “That gives us the ability to collect vast amounts of data and to also actively experiment on what shapes and influences consumer choice and consumer behaviour.”

Still, this can lead to an imbalance of knowledge between platforms and their users, who may not be aware of the user design nudging their choices. Platforms are built to facilitate connection and exchange, and removing friction may actually help to streamline the path to those outcomes. Nevertheless, this power is not to be taken lightly. “It has the potential to be used for good, where platforms can help us to achieve our own goals, to connect better with each other, to learn,” says Elisabeth. “It can also be used, I think, for ill, where there can be a misalignment of incentives between platforms and users.”

Platforms: choice architects that shape our world

Andreas Maaløe Jespersen is Assistant Director at the UK Competition and Market Authority’s Behavioural Hub, and he agrees that it is vital to pay attention to how platforms are developing. “We care about platforms because they are increasingly what is defining markets today, and especially the largest platforms […] now serve as the key instruments in how the market evolves,” he says.

The focus of the Behavioural Hub is to monitor the ‘basic fundamental law of market economics’, i.e. the value exchange of money for services and products. “We think about platforms as something that is driving immense value,” says Andreas. To cultivate dynamic markets, the Behavioural Hub keeps a close watch for potential anti-competition violations that could cause harm. They evaluate how a platform might be enabling or blocking healthy competition, asking:

• Is it making it difficult for others to access the market?
• Is it causing harm to consumers that they can’t see?
• Is its position in the market artificially inflating prices?

Platforms are market-defining businesses with immense societal and economic influence, and this makes them powerful architects of consumer choice.

This sense of scale is partly what drives Andreas’ team and their regulatory work. “A very important part of our work is respecting that there’s a plurality of preferences. Users are not the same, consumers are not similar necessarily,” he notes. “And we want a market that can cater to many, many different types of wants and needs.”

He adds: “It’s our role to then think about how this translates when people interact with platforms […] and then try to extrapolate the consequences on a societal level.”

Platform businesses respond to the behaviours of their users, which can have a self-reinforcing – and often fast-paced – effect on both technology and society. “We need to go a bit beyond and think about how everyday choices and everyday preferences turns into these colossal products in the end, typically because millions and millions of users interact with these platforms,” Andreas urges.

User-centric design: context and iteration

The importance of this responsibility for large platforms and tech companies isn’t lost on Esra Ozkan, UX Lead at Alphabet Inc. For Esra, two major forces drive her approach to thoughtful, user-centric design: context and data-driven iteration.

Contextuality in choice architecture involves holistic consideration of users’ needs and goals. “What is the user trying to accomplish and what role does the choice play in that path? How can we present choices that are relevant to users’ context?” prompts Esra. She adds: “That includes all the way from a user’s tech literacy level, languages they speak, maybe devices they use, and even the level of internet connectivity they have.” This granularity is ultimately in service of better products and user outcomes. “Thinking about our users as as monolithic group of users is detrimental to the solutions. […] All the way from age groups to educational levels and more, [we want to make] sure that there are multiple groups that are being researched and designed for,” she says.

Thanks as well to her data-driven, iterative design philosophy, this creates a two-way process of research and experimentation. “What we think is going to work for users may not work in the end,” acknowledges Esra. For her and her team, iterating on possible solutions and using data to validate assumptions – all while staying focused on user context – is what user-centric design looks like in action.

The tools of the trade

The term ‘choice architecture’ is often associated with defaults and nudges, and Eric has published fascinating research on the impact of defaults on organ donation. However, he points out that design elements like sorting and descriptions can also be highly influential, if more subtle.

“What is first makes a big difference in some settings,” Eric says. “My favourite example is in the US election: it turns out, even for president, whoever’s first on the ballot gets a 1% increase in share.” The same sorting effect has been observed online, as well – for example, in how users choose hotel rooms on a booking site.

Descriptive nuances might go largely unnoticed, but they can also affect consumer behaviour. Eric shares: “A favourite example of mine is describing ground beef as either 70% lean or 30% fat. It’s the same thing, but even those labels can be very powerful.”

Friction: the trade-off between experience and outcome

One of the most pervasive and powerful functions of choice architecture is not only how choices are presented, but the simple fact of when – or whether – they are offered to users. A moment when a user needs to make a choice is referred to as ‘friction’ in the user experience.

Andreas reminds us that this has parallels in offline life. “Friction has been a natural part of our process. We had to go to the store, we had to wait in line. We had to go to the post office to pay our bills.” However, digital environments offer more possibilities to remove these frictions. “We are seeing the evaporation of friction in many instances. […] And I think we are beginning to see some of the negative sides of removing a lot of friction from the markets.”

“UX design often emphasises […] making the experience feel good to the consumer,” says Eric. However, he thinks that fluency and ease of making a choice should be balanced against how accurately that choice will satisfy the user’s needs. “I could feel great about a choice and it turned out not to be very good,” he continues. “So often, there’s this tension between choice quality and the experience of making the choice.”

The data also suggests that our perception of choice quality is often inaccurate. “[People] aren’t as great decision-makers as perhaps economics makes us think they are,” says Eric, citing research that showed as few as 40% of study participants were able to make the choice that would be most advantageous for them when faced with a selection of credit card options. According to Eric: “Choice architecture is a partner in making a better choice.”

Friction in action at Google

As a practitioner, Esra is familiar with the challenge of navigating this tension between fluency and accuracy. “Choice may create friction either because it interrupts the task or because there are too many choices to make,” she explains. “It’s not just Google that [users] are interacting with throughout the day. They have many other ecosystems they go in and out of.” The demands of repeated decision-making could be one factor that inhibits users from making choices that will actually be in their best interests.

In order to not overburden users with information, Google has the task of determining when introducing a choice – and therefore, friction – helps people reach a better outcome. “We have millions of resources that we could provide to users,” Esra reminds us. She has a simple illustration of how this works in practice at Google. Sometimes, users are searching for factual information, such as directions to the nearest hospital. “That’s maybe where we are going to put the choices in the background because the accuracy of the results is what’s most important to the user at that point,” she says. However, she adds, a wide array of choices would be more important to a user looking for used car listings, and the search results should take that need into account.

Conversely, unity across its products is one area where Esra sees an opportunity for Google to further improve its approach to choice architecture. “How do we actually keep choice preferences coherent across these products […] so that users don’t have to make the same choices again and again?” She notes, however, that this aim of offering a streamlined experience for users also needs to be balanced against protecting user privacy.

Browser cookies preferences and misapplied friction

Even well-intentioned choice architecture can have unforeseen, sometimes undesirable results. It’s often a matter of timing. “Friction is going to have the most influence at the very beginning [of a decision],” Eric says. “We often decide how we’re going to make a decision and then we never change that because we’re too busy making a decision to revisit that.”

To illustrate how friction can be applied at the wrong time, Eric brings up an activity that is ubiquitous for any online user: setting cookie preferences. “If I go to one more website and am asked, ‘How do I want to manage my cookies?’ when my goal at that time is to book a flight or to order a new book, that’s creating a friction to what I want to do, and therefore I dismiss [the cookie settings].” It’s also a clear example of why it is vital to have interdisciplinary efforts to globally improve the design of this user experience. Eric adds, “That’s a serious issue [and] it’s because of friction that’s at the wrong time. [Choosing your privacy settings is] an important decision.”

Esra agrees that cookie consensus is an example of misapplied friction. “We all want our privacy to be respected, and I think the regulators had the right intentions and wanted companies to do something about it. But the solution we came up with collectively is far from perfect,” she says. “In the past, we just didn’t click any buttons and we shared information that we might not have wanted to share. And now we just accept cookies – either because it’s the prominent choice or it’s just an annoyance and there’s no time to read and customise what information we share with platforms versus what we don’t.”

The result is that the right choice is not easy to make and the outcome is often not reflective of what users really want. Esra comments: “We still are in the same place from a privacy perspective, I think, except that we have more friction – and I don’t think that’s what companies want, users want or regulators want.”

Responsible design that benefits all

The advantage of bringing together practitioners, policymakers and academics is that discussions can be framed around shared goals and responsibilities. As Eric notes, user-centric design is a fantastic opportunity: “The notion is that people worry about choice architecture as manipulation, but it also can be a very powerful tool for good.”

In order to realise this potential, Esra strongly believes in the need for exchange and dialogue: “Academic scholars, practitioners and regulators have the same end goal in mind but we do see different parts of the picture.” She adds that the practical work of enabling user-centric design is easier when approached collaboratively: “I’m really also keen on finding ways to conduct research collectively that we can’t possibly conduct on our own and finding ways in which we can actually share this research – while protecting intellectual property and participants’ privacy.” It may sound lofty, but Esra says, “I am confident that we can reach a place where we are conducting meaningful research and learning so that we can reach that common goal.”

As a regulator, Andreas also sees the necessity of fostering an environment that is healthy for businesses, as well as users. He adds, “We need to be very strategic because we also recognise that adding friction to a business’ interface is a cost.” And the aim is not to handcuff platforms, but to cultivate a safe and thriving competitive market. “Increasingly, we are seeing platforms who shape large parts of people’s everyday lives: what we buy, what we watch, who we talk to and how. And I think having a stronger sense of responsibility for users is something we’ll see in the future – if not out of the goodness of their heart, then because regulation will evolve and force businesses to be more holistic,” says Andreas.

Still, Andreas believes that regulators should and will increasingly place the onus on businesses for responsible choice architecture: “I think the best platforms for the future are ones who think about consumer and user welfare holistically and who take a more protective view of how consumers interact once they are on a platform.” Rather than regulatory ‘whack-a-mole’, Andreas predicts that regulators will push businesses to take a proactive approach to consumer welfare. For example, users might be asked questions upfront like: ‘How do you want your privacy settings? What type of products do you want to be shown?’

Any regulatory pressure will be joined by pressure from the market, Esra believes. “Whether or not you’re empowering your users in the right way will be almost a selling point.” In her view, accountability is already on the rise. “Better design, a more powerful user experience, effective solutions – and doing it in a responsible manner where we are respecting users’ privacy – are going to be matters of consumer choices.”

In the future, digital platforms will…

Looking to the future, Esra, Eric and Andreas have ideas about how digital platforms will evolve and accomplish. Again, their comments highlight the value building a community of platform experts with diverse backgrounds. To finish the sentence, ‘In the future, digital platforms will…’, all three offer their own unique perspective.

Esra predicts: “Digital platforms will be more complex and require experts from multiple fields to improve.”
Eric’s answer: “In the future, platforms will help improve consumer decision and consumers’ welfare.”
Andreas says: “They will be more responsible.”

To go further

This panel with Professor Eric J. Johnson, Elisabeth Costa, Esra Ozkan and Andreas Maaløe Jespersen was part of the Platform Leaders event organised by Launchworks & Co on the 7th of June 2022 (full list of speakers and agenda). To watch the full event, play the video below.