Coca-Cola has become the latest brand to boycott Facebook as the social network grapples with how to deal with hate content.
After its shares crashed on Friday, Facebook announced a new plan to label all rule-breaking posts, even those from Donald Trump - whom it's been previously loath to touch amid claims that the social network's CEO Mark Zuckerberg and director Peter Thiel have got too close to the US President, who recently dined with the pair.
"There is no place for racism in the world and there is no place for racism on social media," Coca-Cola chairman and CEO James Quincey posted early Sunday NZT.
"The Coca-Cola Company will pause paid advertising on all social media platforms globally for at least 30 days. We will take this time to reassess our advertising policies to determine whether revisions are needed. We also expect greater accountability and transparency from our social media partners."
On Friday, European fast-moving consumer goods giant Unilever, whose stable includes Dove, Rexona, Lynx and Sunsilk, joined a number of companies boycotting Facebook and Facebook-owned Instagram, and the smaller Twitter.
"We have decided that starting now through at least the end of the year, we will not run brand advertising in social media newsfeed platforms Facebook, Instagram and Twitter in the US," Unilever said in a statement.
"Continuing to advertise on these platforms at this time would not add value to people and society."
Earlier this week, clothing chain Eddie Bauer, film distributor Magnolia Pictures and the Ben & Jerry's ice cream brand announced that they would stop advertising on Facebook through July, joining other boycotters including Patagonia, The North Face and REI.
Facebook shares crashed 8.3 per cent on Friday, eliminating $US56b ($NZ87b) from the company's market value and knocking $7.2b off Zuckerberg's personal net worth.
The same day, the social network said it will flag all "newsworthy" posts from politicians that break its rules, including those from President Donald Trump, among other measures.
However, Coca-Cola's walkout announced today indicated the company will have to go further to win back corporate advertisers.
Zuckerberg had previously refused to take action against Trump posts suggesting that mail-in ballots will lead to voter fraud, saying that people deserved to hear unfiltered statements from political leaders. Twitter, by contrast, slapped a "get the facts" label on them.
On Tuesday, Trump tweeted that if protesters attempted to set up an "autonomous zone" in Washington DC they'd be "met with serious force".
Twitter slapped a warning label over the President's message to those planning to exercise their right to peaceful assembly.
"We've placed a public interest notice on this tweet for violating our policy against abusive behaviour, specifically, the presence of a threat of harm against an identifiable group," it said. The tweet could still be read after clicking on the warning.
The label also covered over the tweet, blocking users from being able to view the post, and blocked people from being able to retweet the post. Users can still see the tweet by clicking on it.
It was the fourth time Twitter, had labelled a Trump tweet recently, despite the President signing a rambling executive order that sought to clip social media platform's wings.
Trump shared the same message on social media giant Facebook, where it remained intact and received more than 130,000 likes and has been shared more than 5400 times.
Now, Facebook says it's tightening up. "The policies we're implementing today are designed to address the reality of the challenges our country is facing and how they're showing up across our community," Zuckerberg wrote on his Facebook page announcing the changes yesterday.
Zuckerberg said the social network is taking additional steps to counter election-related misinformation. In particular, the social network will begin adding new labels to all posts about voting that will direct users to authoritative information from state and local election officials.
Facebook now says it is also banning false claims intended to discourage voting, such as stories about federal agents checking the legal status at polling places. The company also said it is increasing its enforcement capacity to remove false claims about local polling conditions in the 72 hours before the US election.
Ethan Zuckerman, director of the Massachusetts Institute of Technology's Center for Civic Media, said the changes are a "reminder of how powerful Facebook may be in terms of spreading disinformation during the upcoming election."
He said the voting labels will depend on how good Facebook's artificial intelligence is at identifying posts to label.
"If every post that mentions voting links, people will start ignoring those links. If they're targeted to posts that say things like 'Police will be checking warrants and unpaid traffic tickets at polls' -- a classic voter suppression disinfo tactic -- and clearly mark posts as disinfo, they might be useful," he said.
But Zuckerman noted that Facebook "has a history of trying hard not to alienate right-leaning users, and given how tightly President Trump has aligned himself with voter-suppressing misinformation, it seems likely that Facebook will err on the side of non-intrusive and ignorable labels, which would minimise the impact of the campaign."
Fake ads persist
Beyond concerns about the environment their genuine ads are appearing, companies like McDonalds, PakNSave owner Foodstuffs have had to suffer fake ads placed by fraudsters running coupon scams (a "McDonalds" effort ran on June 25), while individuals including Richie McCaw and Mike Hosking have had their unauthorised images used in social media ads for dubious get-rich schemes.
With reporting by AP