It’s nearly impossible to not influence you users’ decisions in one way or another. Similarly to the statement “there is no absence of design, just good design and bad design”, there is no absence of decision architecture. Since you can’t help it, you might as well try to influence your users in way that has positive consequences for them.
Paternalism vs. LibertarianismMost of the principles of decision architecture are rooted in behavioral economics. The classical view of the human in an economic context-the homo economicus-proposes that humans always make the most sensible decisions, given they possess all the information needed to decide. However this idea of man has since been replaced by a more realistic view.
This realistic human image accepts that humans aren’t usually able to make ideal decisions for various reasons ranging from a lack of self control to simply being overwhelmed by possible options. This is where paternalism comes into play. Paternalism is the practice of limiting a person’s or group’s liberty, options or choices with the intent of increasing their overall good. Examples for paternalistic behavior is limiting children’s access to sweets, putting age restrictions on movies, not allowing users to submit forms unless they are completely filled and laws that require you to wear a seatbelt.
While paternalism is well intended, it also limits people’s freedom and is therefore often critiqued. Especially by libertarianism, which tries to protect people’s right to choose freely. This puts the interaction designer on the spot making a decision between paternalism and libertarianism. How can they bridge the gap between users’ reality and the ideal UX? The solution is using positive reinforcement and indirect suggestion to influence users’ behavior and decision-making and to gently nudge users towards the intended behavior while always giving them the freedom to choose otherwise.
Guiding users in a certain direction while not using restrictions or laws and always giving them the option to decide otherwise is called libertarian paternalism or “nudge”.
How to nudgeIf the decision exceeds an individual’s mental capacity or the person deciding is under pressure or has time-constraints, their decision will rely heavily on heuristics and environmental influences. Therefore nudges are especially effective when decisions need to be made in situations that are overly complex or overwhelming. Nudges should be used to encourage positive behaviors that benefit the decision-maker. An often used example is putting healthy food like fruit and vegetables in places where people will see them when they first enter a cafeteria. That way they are more likely to show healthy behavior while also being able to choose less healthy food which is placed somewhere after the healthier choices. To make sure that you design nudges in an ethical way:
- Nudges have to be transparent and should never confuse users
- Deciding against the behaviour encouraged by a nudge should be as simple as possible to ensure people’s freedom of choice
- People should be able to assume that the proposed behavior is beneficial to them or the general public
Nudges rely on relatively simple behavioral principles. For example people tend to accept the default. So if you pre-select an option for your users, you can assume that they are more likely to simply accept it. If you increase the ease of use for a certain kind of behavior, people are more likely to behave that way. You can use reminders for intended behavior if it can be assumed that users will forget about something. However don’t bother users with constant reminders that they don’t want or need. Lastly you can frame the decision in a way that encourages the intended behavior.
When not to nudgeEarly in this article I stated that decision architecture tries to “gently nudge […] users towards the intended behavior”. If you think this sounds eerily like manipulation you’re not alone. Libertarian paternalism proposes that it is both possible and legitimate to affect people’s behavior while also respecting their freedom of choice. Still the “paternalism” part of libertarian paternalism implies that the users you’re trying to (positively) nudge, don’t always want to make the “beneficial choice”. Quite the contrary: sometimes they are pushed towards a certain choice against their will.
As the person influencing your users’ decisions you can have the best intentions but still need to accept that you never truly know which decision really is in line with your users’ intentions. So when you nudge, do so gently and give users the option to push back. Don’t be pushy or your well intended nudging risks to slip off in the direction of dark UX patterns. Scenarios to avoid include:
- Using nudges in common scenarios or cases where users face simple choices and are provided with more than enough feedback
- Nudging users towards decisions that don’t benefit them
- Repeatedly asking the same question and not providing the user with an “don’t show this dialogue again” option
- Implying that by choosing one option, users are missing something when they are actually simply making a choice
- Repeatedly asking for confirmation, implying that the user has made a choice that is wrong or dangerous
- Setting bad defaults that no one would ever choose voluntarily
- Hiding things that should be obvious or at least noticeable
- Putting things they didn’t ask for into your customers‘ shopping carts
- Making people chose an option they would never really want to chose
To sum it up: you should generally avoid to design experiences in a way that makes users decide in a way they didn’t want to. Instead set defaults to the most helpful option, make downloads accessible without unnecessarily gating and only email people what you told them you would.