On March 2, 2018, United Airlines announced a new company initiative. It planned to stop giving employees quarterly performance bonuses of up to $375. Instead, qualifying employees would be entered into a quarterly lottery for cash, cars, vacations, and one grand prize of $100,000. To qualify, employees would need perfect attendance, and the company would have to hit its targets. According to a New York Times article by Christina Caron, those receiving bonuses would drop from roughly 80,000 to 1,361, with approximately  61 big winners.

Within days of the announcement, United Airlines President Scott Kirby announced they would be “pressing the pause button” on the lottery bonus plan, based on outraged employee response.

It’s possible the idea came from a place of good will. But, what we know for sure is that they stepped into a fresh mess of unintended consequences.

Much of this situation was both predictable and preventable.

Here’s what they should have considered.

  • Humans are hard-wired for fairness. Let’s say you’ve been told you must split $100 with someone you don’t know and will never meet. The catch is, you both must agree on the split. If either of you rejects it, you both forfeit the money. Your mystery partner has offered you $10 and will keep the remaining $90. Will you take it or forgo it all? Researchers have replicated this “Ultimatum Game” since 1982 and found most people naturally offer fair deals, and reject unfair ones even if they benefit from them.
  • Small losses hurt and focus our attention. How much do you care about seven cents? Not much, right? Yet, in 2017, the idea of paying seven cents for a grocery bag convinced 33% of Chicagoans to bring their own bags. Various counties in California have experienced similar results since instituting a ten-cent fee in 2012. A nominal cost focused people’s attention on an issue and changed group behavior.
  • Predictability matters. United Airlines employees expected to receive $375 a quarter. To disrupt that is to violate an expectation, a social contract. And it’s a change – it’s new. Because we are wired for survival, we unconsciously look for anything unusual. Any disruption to “normal” is by nature threatening.

United tried to sell this as “exciting,” but everything its employees felt in their bones told them otherwise. Compensation design issues aside, here’s what to consider when implementing a massive change like United’s.

  • Start with the bad news. We so prefer the predictable that we will stay in a bad situation until it is completely broken. This tendency keeps us in toxic jobs, expired relationships, and unreliable vehicles. So, resist the temptation to talk about benefits first. You must convince people that the current situation is harmful, dangerous, or impossible to sustain.
  • Frame the issue properly. Clearly define the problem you are trying to solve and communicate it honestly. Was United’s problem that the overall benefits package was too expensive to sustain? Was it that small bonuses appear to have little impact? Most people can take terrible news if they have a chance to understand the situation, the implications, and how it impacts them directly.
  • Establish the right comparisons. Comparisons are how we define value – why the outrageous $1,200 airfare you looked up two minutes ago (your anchor) is suddenly a screaming deal once you found the same flight for $2,000. In United’s case, management anchored against big prizes and big savings, while employees anchored against their current quarterly bonus. If management had identified what employees used as their anchor and aligned on that, they would have addressed the inherent fairness issue. Ask employees what’s on the grapevine about your initiative, then ask specifically what people are comparing it to. Make sure your point of reference is the same as your employees’.
  • Launch at least three times. Borrow from Everett Rogers’ research on how innovation spreads. Your initial pilot should be between 2.5% – 13.5% of your intended population. Identify employees who are naturally enthusiastic about new ideas and involved in designing them. These positive, leading-edge people will provide insight on how to adjust the program while maintaining your intended objective. Then pilot again with 34% of your intended group but pick only those who are generally positive about change. They will continue to offer constructive feedback and positively feed the rumor mill. This makes the change safer for people who are more hesitant to sign on. Once these two pilots are complete, you have a more controlled environment for a full launch.
  • Make the first step excellent. Research shows first impressions color our assessment of everything that follows. Manage that impression vigilantly – if your new IT system is complex, focus on rolling out only the simple sign-on. If your new compensation model is revolutionary, implement the spot bonus first. Design small, painless steps that allow people to gain familiarity with your direction. As people see themselves succeeding in the new system, they will come on board.

Every leader faces the threat of unintended consequences when they deploy something new. Fortunately, we have access to an unprecedented volume of research on human behavior. Look to that work to make decisions that affect your organization and your people. Give it the same emphasis as you would the technical aspects of your project. This will help you avoid false starts and mitigate those uncomfortable outcomes.