X Spent Two Years Winning Back Trust. Grok Lost It in a Day

On May 12, 2023, Elon Musk named Linda Yaccarino, NBCUniversal's global ad chief, as CEO of Twitter, specifically to bring advertisers back after the acquisition rattled them. On Jul 23, 2023, Twitter became X, and major advertisers paused spending over brand-safety concerns anyway. Yaccarino spent two years courting them back. On Jul 8, 2025, X's Grok chatbot posted antisemitic content that drew widespread public condemnation. On Jul 9, 2025, Yaccarino announced her resignation.

After two incredible years, I've decided to step down as CEO of X.

Two years of relationship-building evaporated inside a single news cycle. That's not a coincidence of timing. It's how the people who control ad budgets actually make decisions when something goes wrong in public.

The principle: loss aversion, backed by social proof

Loss aversion means people weigh the pain of a potential loss more heavily than the pleasure of an equivalent gain, which makes avoiding a bad outcome a stronger motivator than chasing a good one. An advertiser who pulled spend from X after the Grok incident wasn't calculating the upside they'd give up. They were calculating what they stood to lose if their own brand got linked to the platform's failure.

Social proof compounds that calculation. People look to the actions and approval of others to decide what is correct or worth doing, especially under uncertainty, and brand-safety decisions are made under exactly that kind of uncertainty. No single marketing team has full visibility into how bad a platform incident really is or how long it will last. So they watch what other advertisers do. One or two visible pauses can become the signal that tells everyone else pausing is the safe, correct move, regardless of whether any of them individually believe the platform is now unsafe.

The gap between the rebuild and the reaction

Yaccarino's hiring was framed as a solution to a trust problem, and for two years, the strategy worked on its own terms: advertisers came back. What that strategy didn't account for was how little of that trust was actually load-bearing. Two years of relationship management had addressed the original brand-safety concern from the 2023 rebrand, but it hadn't changed the structural reality that a single platform incident could reopen the same question for every advertiser at once.

The expectation, reasonable on its face, was that sustained outreach and relationship work would make the next controversy survivable. What happened instead was that the next controversy didn't need to be worse than the first one. It just needed to happen on a platform where advertisers already had a recent memory of pulling out, a recent memory of other advertisers doing the same thing, and a CEO role whose whole mandate was proof that the risk had been real enough to require fixing.

That's the gap. Trust rebuilt through persuasion doesn't behave the same way as trust that was never broken. It's thinner, and it's held by more people watching each other for the first sign that pulling out again is acceptable.

What a behavior-first process would have surfaced

None of this is a hindsight problem unique to X. It's the standard failure mode of planning a campaign or a brand relationship around what you want your audience to think, instead of what that audience is actually primed to do under pressure. Advertisers weren't a monolith reacting to Grok in isolation. They were a network of decision-makers each watching for permission to act on a loss they were already primed to fear.

This is the kind of dynamic DayClerk's audience simulation step is built to surface before any copy or creative ships: it looks at a brief and models the behavior paths, friction points, and drop-off risks specific to a segment, rather than assuming one message lands the same way for every audience regardless of what they're watching each other do. A brand managing a sensitive relationship, whether with advertisers, partners, or a wary customer segment, benefits from seeing where the friction already lives before committing a message to it.

The takeaway

If part of your audience makes decisions by watching what others in their position do, don't plan your campaign or your crisis response around individual persuasion alone. Map out what happens if a few of them move first. Loss aversion plus social proof doesn't need a majority to tip a room. It needs a handful of visible exits and everyone else deciding that following them is the safer bet.

Note: AI was used to help write, format and edit this article. Reviewed by the DayClerk team.