Designed to Nudge: a Prize-Winning Idea

By November 2, 2017Embedded BI
mama and baby elephant

Wouldn’t it be nice if we could get users to take the correct action at the right time? If we design our applications to “nudge” them into the right behavior, it’s possible.

Last month the Nobel Prize in Economics was given to one of the originators of Nudge theory, behavioral economist and author Richard Thaler. Traditional economic theory is premised on the idea that consumers are rational and make decisions based on informed self-interest. Thaler’s studies showed that not to be the case; that actually, when it comes to decision-making, we tend to be highly irrational.

Izenda’s mission is to empower users to make better, data-informed business decisions. Thaler’s studies suggest that one way to do this is with a “nudge”. A nudge is a design element that subtly steers people into making better decisions, while requiring minimal effort to do so. Nudges can take the form of physical things like traffic signage or store design, but they can also be part of an application’s UX and processes.

Cities and states have started to use nudges to guide citizens to making better decisions. In New Mexico, 2015 enrollees for unemployment insurance were nudged using messages personalized during the application process. The state wanted to cut down on fraudulent claims. When the system associated fraud-related risks with a claimant’s answers, it triggered personalized pop-up messages to gather more information.

New Mexico beat the national average for reduction in overpayments between 2015 and 2016 with a 35% reduction in fraud, or $1.9 million dollars annually. Those little nudges had a big effect.

How Can Analytics Support Nudge Behavior?

In a world where we are faced with hundreds of tiny decisions each day, it’s important to design to help reduce cognitive overload. Nudges can be essential for this. They can take different forms, many of which will be recognizable by architects and UX designers:

Choice architecture: Displaying the better of two options clearly, for example, by putting fruit and healthy snacks on a higher shelf than junk food in a school lunchroom. Embedded Analytics use: displaying the most important visualization or metric, such as a gauge with relevant KPIs, above the fold and in an eye-catching color. Other data could be displayed below the fold with more subtle coloration and potentially in grid format.

Reminders: Prompting users to carry out a simple action, like an ATM pinging you to remove your card. Embedded Analytics use: an email alert that automatically triggers when a condition is met or a message in a pop-up or in an alerts section on the top of a page following log in.

Default settings: Developers tend to set “off” as a default. A nudge would be to default to opt all employees into a 401K. Embedded Analytics use: automatically subscribing users to appropriate reports and alerts that are relevant to their role. This could serve to nudge people to review vital but underutilized reports and dashboards or to help them view the most critical information.

Status displays: Most of us are by nature driven to compete with others or to complete an action when we are close to a goal. An example of a nudge is a fitness application dashboard that display metrics like “you’ve completed x% towards your goal”. Embedded Analytics use: dashboard KPIs that display user activity in relation to other users or as a percentage towards completion of a goal.

Nudge Lightly

There’s a fine line between helping users and irritating them. A nudge should be a gentle push, not a demand. It has to be effortless and minimally intrusive. Even for something as simple as setting a default value, it’s important to weigh the benefits of a nudge against users’ trust and sense of autonomy when using your application.

Marketing (even without a Nobel prize…yet) can guide those looking to provide the right magnitude of nudge. Relevance to the user is key, and cohesiveness with user workflows essential. While alerts are designed to be jarring, nudging users to change their behavior in other ways requires a more subtle approach. Since most BI tools are typically adopted by less than 30% of users, using nudges as subtle reminders at key points in workflows is one way to ensure users see the value of the analytics embedded in their application. A few best practices:

  • Position information and use color so that it doesn’t overwhelm the user. If everything is a crisis, the user will become numb to the information overload.
  • Make it clear where to go to get more information. Providing dashboard and report widgets throughout the workflow is key, but making it clear that they can be drilled down and through is also essential.
  • Make analytics an integral part of the application, not an iFramed experience that interrupts their workflow and reinforces that they are in a separate application.
  • Empower the user to follow through. It is not worth nudging someone if the application leaves them unable to act on the nudge.
download the ebook Overcoming the 7 Challenges to Embedding BI in Applications

Leave a Reply