Crowdsourcing Mistakes

While the practice of conducting A/B tests—that is, gathering data on which of two options works better—dates as far back as Frederick Taylor, the ability to do A/B testing fast and cheap was how Amazon and Google were able to optimize their services so rapidly. Finding innovations through experiments at tech companies is now standard practice.

So, why don’t most organizations do this when it comes to managing people?

For the past several years I have served on an advisory board for Express Scripts Inc., the largest pharmacy-benefits manager in the United States. The board is made up mostly of entrepreneurial academics who suggest ways that Express Scripts can improve the services it offers.

But the real secret that we’ve helped to unlock: The executive team now views every decision as a data collection opportunity.

Rather than argue that some policy is “right” or “wrong,” they instead identify the goal they’re seeking and design a small-scale experiment to see if the policy moves them toward this objective. Notably, this is now done not only for production and service delivery, but for internal policies that affect the workplace environment—policies that are meant to spur higher productivity, greater employee engagement and enhanced interpersonal trust.

For example, companies such as Netflix and IBM have eliminated vacation policies and now let employees choose when and how often they go on holiday. If your company is considering this, you might try a no-vacation-policy experiment for a year at a one location and track the effects on productivity, morale and healthcare usage to assess whether you should roll out this arrangement companywide.

Running experiments like this requires a change in mindset. Indeed, before executives can view policy changes as experiments, they have to embrace the idea of “smart mistakes.” Trying and failing is an opportunity to learn. Even better: When employees are engaged in this endeavor, they are empowered to create rather than punished for failing.

Think of it as the crowdsourcing of mistakes in order to find improvements.

You can add even more power to the concept by turning your experiments into a corporate custom. You might give a prize for the best idea, or wear funny hats during the meeting where you pick A or B. The key is to repeat whatever you’re doing every time out. My lab has shown that participating in rituals of all types causes the brain to release oxytocin, stimulating trust and cooperation.

Peter Drucker certainly understood the wisdom of learning from failure. “The better a man is,” he wrote, “the more mistakes he will make for the more things he will try.”

At your next all-hands meeting, trying soliciting ideas for A/B experiments.  Then, run one and see what your organization learns.

Paul Zak is the director of the Center for Neuroeconomics Studies at Claremont Graduate University and the author of The Moral Molecule.