How A 1930’s Harvard Student Laid The Groundwork For Facebook Likes

3 min read

This is a bonus excerpt from our upcoming eBook Digital Behavior Design.
Click HERE to reserve a FREE Copy.


Despite All Your [Internet] Rage,

You Are Still Just A Rat In A Cage


The Skinner Box

While completing his graduate coursework at Harvard in the 1930’s, world-renowned neuroscientist and behaviorist B.F. Skinner set out to learn how behavior could be shaped and defined.  Working with rats in a box that Skinner himself designed he observed that there were factors that could increase or decrease the frequency of a behavior.  His primary observation was around positive reinforcement which rewarded the subject for the behavior Skinner wanted them to exhibit. Skinner found that actions which contained positive rewards that were pleasurable or delightful were performed at a higher frequency while actions that received no reward or were painful were more likely to be abandoned or avoided.  By giving his subjects (rats) rewards at the optimal time he could increase the likelihood that would come back and do the behavior more.  

The Skinner Box in Our Pocket

After Skinner, designers and technologists became obsessed with finding new and innovative ways to create behaviors around every day actions.  While not inherently obvious many of the behaviors you exhibit today around commerce are derived from Skinner’s early findings. Scanning your rewards card to receive points at a drugstore?  Congrats, like Skinner’s rats pushing a button you’re presenting that card because of past rewards. Choosing to workout with a personal trainer who can sense when you are struggling and need that extra High Five to push through?  Yep, still a rat.

No greater invention in the modern era has led to more behavior change than the iPhone.  Smartphones have changed behaviors around how we travel, eat, exercise, communicate, love, date, and play.  We use our phones to help us run marathons, stick to strict diet plans, pay back loans, and communicate more efficiently with our friends.  Of course, after completing these actions we are almost always rewarded with a badge, a like, a notification, or soft vibration. 

Take for example Fitbit which drives behavior change by rewarding you when you complete your step goal.  If you hit your goal you are always shown a badge or notification letting you know you did something good.  This notification serves two purposes.  The first is to alert you that you have completed the desired behavior.  The second is to make you feel good about completing the behavior. The better timed (and more surprising) the reward the more it changes your behavior.  The two go hand in hand and help complete a habit loop (just like the original skinner box) of action -> feedback -> reward.  


The Skinner Box of The Future

At Boundless Mind we’ve built a Skinner box more powerful than anything B.F. himself could have imagined in his wildest dreams.  Using the knowledge we gained as Neuroscience PhD students plus the latest in artificial intelligence and machine learning we are able to tell exactly how and when each individual user of an app needs to be rewarded to drive the preferred behavior.  Our AI learns from the past and present behavior of each user and then accurately predicts when the user will need to be rewarded in the future. To put in simply, we have built a self-training system that learns in real time and delivers rewards at the time the user will find it most unexpected. Behavior design isn’t luck or magic, it’s science!


Sound too good to be true?  Below is a real graph of a Boundless Mind customer operating as a lifestyle productivity app (let’s call the app “TodaysToDo”) that wanted to reward users for the action “check off item”. TodaysToDo integrated with Boundless Mind and decided that they wanted to show a simple confetti explosion when the action “check off item” was completed.  Our AI system observed their users behavior, decided which user and at what frequency they should be rewarded for the requested action, and then delivered the reward exactly when the individual user needed to see it most.  Here is what happened:

The green line represents users who were not exposed to the perfectly timed rewards from the Boundless AI and the blue line represents those that were.  The gap between the two lines shows a 30% boost in the target behavior “check off item” which in turn means higher engagement, longer retention, and a healthy boost in revenue for TodaysToDo.  The Boundless AI knew exactly which users needed that extra nudge after completing one of their daily tasks and rewarded them for doing so.  A simple action with a simple reward led to a massive boost in revenue.

Behavior change isn’t luck, it’s science.  At Boundless we’ve figured out how to drive positive behavior change with minimal engineering resources.  Whether you are a team of two in a garage building the next great mobile game or a product manager at a Fortune 100 company Boundless Mind can help your organization get the most out of your users.  Cut Churn, Hack Behavior, Be Boundless!


Matt Mayberry