Why Zuckerberg reversed course: it wasn’t just the boycotts

Boycotts rarely work. But in the face of advertiser boycotts, Facebook just did a U-turn on its policy on labelling or blocking presidential posts. The cause and effect here is a lot more complex than it appears.

As recently as the beginning of June, Facebook CEO Mark Zuckerberg said the right decision on presidential posts, no matter how incendiary, was to “leave them up.” Now he says Facebook will hide or block content that is hateful or could could harm voting, regardless of who posts it.

What changed?

Boycotts were the start of this shift, but not the end of it

The obvious answer is “boycotts.” Leaders of groups like the NAACP and the Anti-Defamation League called for a suspension of advertising on Facebook. Since then, 90 marketers, including Coca-cola, Starbucks, Verizon, REI, Honda, Patagonia, and Unilever have pulled ads for July.

But I don’t think the boycotts are what caused the change — at least not alone. First of all, Facebook has over 7 million advertisers. Facebook ad spending is likely down, just because ad spending everywhere is down, but the temporary loss of large advertisers would not cripple Facebook, because there are so many small advertisers to take up the slack. I don’t see moviemakers and TV shows stopping their Facebook ads; they need the visibility.

Zuckerberg doesn’t sway with the wind, and for good reason. Once he telegraphs that he is vulnerable to influence from boycotts, there will be many more boycotts. And this piece in the New York Times implies that there is a wink-wink deal between Facebook and the Trump administration: the government agrees not take antitrust action, and Facebook leaves Trump’s posts up. Early Facebook investor Roger McNamee said, “I believe they have a deal,” but added that it was “probably implied rather than explicit.”

So what changed?

Three things.

First, in response to the ad boycott, the Facebook stock price dropped 8%. This isn’t sufficient by itself, but when Facebook’s shares go down, it become harder to retain and reward employees. That’s a problem

Second, Facebook employees are revolting. It’s clear that Facebook finally has an internal uprising that is unlikely to go away. Here’s Zuckerberg trying — and failing — to tapdance around the problem at a company meeting.

Third, Trump appears more likely to lose the election. A Democratic administration is more likely to pursue antitrust action against Facebook (Attorney General Elizabeth Warren, anyone?). If Trump can win the election, that’s better for Facebook. But if the perception at Facebook is that he can no longer win, there’s no longer any reason to help him.

There’s a reason I haven’t cited: a moral reaction to Trump’s actual behavior. Reacting to that would imply that Facebook has a moral line that it cannot be pushed past. I’ve seen no evidence of that. Their fealty is to the algorithm.

When do boycotts work?

I’ve said that boycotts are madness. If you hate Jeff Bezos and Amazon, do you stop reading the Washington Post, which he owns? The history of boycotts shows that people lose interest, they don’t make much impact, and they don’t often matter.

There are three groups that boycotts affect:

1. Customers.
2. Advertisers or other partners.
3. Employees.

I believe most boycotts fail because they engage only one group: consumers. Most companies can withstand a 2% drop in customers, especially if they feel that changing positions would cause people with an opposite opinion to boycott them in retaliation. As Kellogg School management professor Brayden King said, “The typical boycott doesn’t have much impact on sales revenue.”

The calculation is: let’s not get involved in politics.

But if a boycott can mess up relationships with customers, advertisers, and employees, then the company has a problem. You can’t do business if you’re tangled up with problems at every step and your employees are complaining and bolting.

This should inform the strategy of those who want to create change with boycotts. They should:

• Create a concrete and realistic demand for change (like asking Facebook to label hateful, distorted, or violence-inciting posts from the President).
• Reach out beyond consumers to partners, advertisers, suppliers, and others who do business with the targeted company. Request specific actions from them.
• Reach out to and incite revolt or calls for change from employees.

A boycott movement that can accomplish those goals is more than symbolic. It has a good chance of creating actual change. That matters a lot more than a few people refusing to buy a brand or use a product, then slowly drifting back as nothing actually changes.