Last month, Stanford’s David F. Larcker and Brian Tayan published 2018 CEO Activism Survey, a U.S. survey (included at the end of this post) of 3,544 people across age, gender and political leanings to ask about CEO activism. The findings in the survey echo others in the past couple of years. Almost two-thirds believe CEOs should be CEO activists, but millennials have very different expectations from brands and CEOs than older cohorts. 71% of millennials expect “CEOs of large companies to use their position and potential influence to advocate on behalf of social, environmental, or political issues that they care about personally”, versus 46% of baby boomers.

Interestingly, Larcker and Tayan show that there’s also a big schism between Republicans and Democrats – 72% of Democrats agree, versus 57% of Republicans.

Both statistics tell both brand managers and CEOs something that should already be obvious – when deciding if, when and how a CEO should speak up, it’s critical to know who your target customers are, and the relative lifetime values of various subgroups. In a world where staying silent is no longer equated as staying neutral, CEOs are being increasingly put in a position where they have to speak up or take an action, and it’s a given that some people will agree and some will disagree. Knowing whether your alienating an occasional shopper with no brand loyalty, or a young loyalist with a potential 40-year relationship with your brand can make the decision easier. Not knowing can be fatal.

Nike’s decision to run the Colin Kaepernick ad campaign is the most obvious example of this. Nike was counting on the loyalty of its core customers’ and brand advocates’ support of Kaepernick to be solidified, knowing it would likely alienate another group of people. I’m willing to bet the company knew exactly who fell into the latter camp – people who buy white running shoes and only at the outlet mall and don’t have 40 years of purchasing power ahead of them.

While Larcker and Tayan reinforce the importance of CEO activism, the gap in academic research that still remains is a longitudinal study of what shoppers actually do in the face of that activism. It’s one thing to say they will switch toothpaste or banks because the CEO pleased or offended them. It’s quite another to actually follow through and change long held shopping habits or resist a sale. Some shopping habits are stickier than others, too.

As we know from the 2016 U.S. presidential election, even voting outcomes – a one-time action – can differ widely from advanced polling. Not everyone who says they are going to vote actually does, just as not everyone who says they’ll change brands does. To truly understand the impact of CEO activism on business outcomes, more data is needed. Certainly Nike results over the next few quarters and the rise in its stock price may hint at the answer, but it’s the long game that really matters.

Here’s the Larcker and Tayan paper.