Inverse Variable Rewards

We can fight smartphone addiction using the same patterns that designers leveraged to make products addictive in the first place. In this article I’ll describe one approach, which I call Inverse Variable Rewards.

Variable rewards are the foundation for a significant number of ‘addictive’ consumer products, used by designers in an effort to ensnare user attention. They seek this attention not necessarily because they think it is in the user’s best interest, but so that they can then monetize it via advertisements, which in turn enables the product to survive and thrive.

Background: What are Variable Rewards?

To understand variable rewards imagine a slot machine. You put in a coin. You pull the lever. Do the three shapes all match? Nope? OK, pull again. How about this time?

Sometimes pulling the lever gives you a reward. Sometimes nothing. (Both times, it cost you 25 cents.) You have no real control over whether you get a reward or not, you only have control over whether or not you pay to play. Looking at the data, slot machines make a lot of money.

It turns out we animals are particularly responsive to rewards that are offered on a variable, non-fixed schedule. This was identified as far back as the 1930s in well-known research by the psychologist B. F. Skinner that involved putting animals in boxes with reward mechanisms. The anticipation of getting the reward (whether or not we actually get one) increases the dopamine levels in our brains, which compels us to keep doing the thing that got us a reward before. Unpredictable rewards based upon a variable schedule result in higher levels of dopamine. Variable rewards also tend to cause the desired behaviors to be repeated more often.

B. F. Skinner & Rat — photo from Life Magazine ca. 1960

Variable Rewards in the Attention Economy

Take a closer look at your phone or your social media. Surreptitious variable rewards systems are practically ubiquitous; there are a number of popular and oft-copied design patterns that use them. Here are a few examples:

• Scrolling down on Facebook’s bottomless newsfeed / Pull-to-Refresh on Twitter: What story will come up next; will it be something funny? Or interesting?
• Swiping on Tinder / any dating app: Will the next person be the one?
• Posting or commenting on social: How many likes/responses will it get me?
• Checking your messages or email or notifications: Has my friend written me back, yet?
• A Notification badge on your iPhone app grid. What’s new inside?

The above features all share one trait: they have no stopping point and thus distract in perpetuity. This is an intentional facet of their design. On the internet, you’re not always paying with money, but you do always pay, with something more scarce and far more precious: your attention.

Variable rewards became a popular marketing and product design concept back in the Web 2.0 era thanks to marketers like Nir Eyal. Attention-hogging consumer services like Facebook, Instagram, and Twitter achieved scale and market dominance through the addictive variable rewards that they designed into their core usage patterns:

Figure 1. Variable rewards as commonly used by consumer websites and apps to capture attention

Each of these services had some value, but they didn’t scale on value the way that Google or Apple initially did. They scaled on dopamine hooks. Today a lot of users are questioning how essential all of these attention traps really are, but regardless, most (if not all) consumer platforms use variable rewards to ‘hook’ your attention.

In fact, attention is the default currency for online variable reward business models. This is a bad situation because it can lead to distracted, stressed, or depressed users. Variable rewards are unhealthy and unsustainable when the user is asked to pay with their attention. To use them this way is unethical. Surreptitious at best; nefarious at worst.

Inverse Variable Rewards

Here’s the good news: variable rewards are incredibly powerful but are not inherently evil. We can use them to ‘give back’ a users attention just as much as they’ve been used to capture attention. By inverting the ‘payment model’ of variable rewards, we can leverage them as a tool for healthy behavior change.

Instead of asking users to pay with attention, ask them to pay with abstinence from use. Instead of being rewarded for looking at, tapping on, or posting something, users could be rewarded for not doing those things.

Does that sound impossible to you? Crazy designer talk? Just stupid? Consider a few examples:

• Starting a new routine. What if your phone enabled you to describe thoughtful intentions and used variable rewards to reinforced your intentional disuse during designated times? Maybe you want to get better at going to the gym at 7am. If you make it to the gym and your phone stays off, you have a chance at getting a reward. If you’re sitting in bed scrolling through Twitter at 7:07AM, you don’t. The reward could be a gym credit, or really anything from bonus emoticons to a few dollars off your next phone/data usage bill, or credits to spend in the app store, or similar.
• Safer Driving. What if your auto insurer offered variable rewards for keeping your phone closed while driving? Imagine arriving at your destination after a distraction-free drive and checking to see if you got a reward — perhaps an insurance credit, or over time a reduction in your insurance rate to reflect the reduced risk of insuring you.
• More balanced media consumption. Imagine that the Facebook newsfeed had a bottom. Users would discover this bottom after some reasonable amount of scrolling or swiping, and then they’d see a message that more content would be loading soon. Facebook could implement a variable reward into this pattern: the possibility of getting a reward when the new content gets loaded. The longer the user waits for more content, the more likely they’d be to get the reward.

Applied in the right situation, Inverse Variable Rewards don’t seem ludicrous at all; in fact, they seem commonsense. They are a way to create balance in a usage system and to make it sustainable. In contrast, a phone nagging for me to look at it while I’m operating a 2-ton vehicle at speed is inherently ludicrous. Apple and Google don’t benefit if I die prematurely in a car crash.

Figure 2. Balancing a system using inverse variable rewards

The above examples demonstrate the simple truth that context matters. Your phone and apps should reward disengagement not all the time, but in a contextual way in order to promote balance, health, and safety. Similarly, social media platforms like Facebook might start incorporating Inverse Variable Rewards to curb usage behaviors that they’ve identified as abusive ones. With all of the usage data they collect on their individual users, they certainly have robust engagement metrics and user profiling. Perhaps they could begin by incenting disengagement from the most addicted 5–10% of their users.

Why on earth would an attention capitalist / ad platform like Facebook do this? Because it helps their platform become more sustainable and less exploitative over time. That probably sounds pretty good to them right about now.