There is a specific kind of paralysis that descends on brand teams when the internet turns on them. Someone walks in with a screenshot. Legal is on the call, muted. Someone from comms has already drafted three statements. The CMO is asking whether this is a "wait it out" or "get ahead of it" situation, as if those are the only two options.

What nobody says out loud, but everybody is thinking: just make it go away. Sand down whatever edge caused this. Go quiet until the algorithm finds something else to be angry about.

This is the instinct that has quietly destroyed more brand equity than any boycott ever has.

The brands that capitulate to it — that spend their crisis meetings hunting for the safest possible middle ground — don’t emerge from controversies stronger. They emerge smaller. A brand-shaped object that could belong to anyone and therefore belongs to no one. Cautious enough to avoid a bad headline. Not interesting enough to make anyone care.

The brands that survive, and the ones that actually benefit from controversy, did something different before the meeting ever happened. They knew who they were for. They’d made the customer segment trade-off explicitly, not by accident. They’d built a position so clear that most boycotts were, structurally, irrelevant — because the people with the loudest grievances were never their customers anyway.

The data on this is not ambiguous. And it should change how you run that meeting.

July 2020. Robert Unanue, CEO of Goya Foods, stood in the White House Rose Garden and praised Donald Trump. Within hours the hashtag was everywhere. Boycott posts accounted for 75% of all social media chatter about the brand. Over 1,500 media stories ran with “Goya” in the headline. Alexandria Ocasio-Cortez posted her homemade arepas recipe as a pointed alternative. The brand appeared finished.

Actual sales impact: a 22% increase.

Cornell researcher Jūra Liaukonytė tracked the household purchase data and found the boycott was real — purchases dropped modestly in Democratic-leaning counties for about eight weeks. But the buycott was larger, faster, and came from a pool with no ceiling.

The math is brutal. Only 7% of American households already bought Goya products. That’s your maximum boycott pool — people who actually have something to stop buying. The buycott pool? Unlimited. Any shopper in any grocery store could grab a can of beans they’d never bought before. And they did. Republican-leaning counties saw massive first-time purchases from people who had never thought about Goya in their lives.

Three weeks later, the whole thing evaporated. Buycotters went back to their regular brands. Most boycotters did too. Core Latino customers — Goya’s actual base — barely changed their behavior at all. They’d been buying Goya for decades and a Rose Garden appearance wasn’t going to rearrange their pantry.

Seventy-five percent boycott sentiment online. Twenty-two percent sales increase in reality. If you’re making brand decisions based on social media volume, you’re not reading the market — you’re reading a very loud, very unrepresentative sample of it.

Monroe Friedman spent years cataloguing consumer boycotts and found that only 26.7% of the 90 he studied from 1970 to 1980 achieved any level of success. Northwestern’s Brayden King analyzed 188 boycotts between 1990 and 2005 and found the same stubborn pattern: the typical boycott doesn’t impact sales revenue. Yale’s research puts the current success rate somewhere between 25 and 40% — but that only counts boycotts significant enough to study. The thousands that fizzled in a week don’t make the dataset.

The timeline is almost metronomic. Peak outrage hits in week one. Social volume crests, media picks it up, the brand’s PR team starts earning their retainer. By week three, attention has fragmented onto whatever crisis arrived next. By week eight, most brands are functionally back to baseline.

The University of North Florida, studying why, was blunt: the majority of boycotts become “too inconvenient and expensive to maintain.”

Convenience is the thing boycott advocates never want to talk about. Amazon has faced organised boycott campaigns since 2012 covering tax avoidance, warehouse conditions, environmental impact, and its relationships with law enforcement. It keeps growing. Boycotting Amazon means rearranging your entire domestic infrastructure — returns, deliveries, streaming, cloud storage, the Kindle your partner uses. People will say they’re boycotting Amazon. Most of them mean they’re annoyed at Amazon.

Facebook’s 2020 ad boycott drew over 1,000 participating companies, generated enormous press, and produced no measurable revenue impact. Chick-fil-A has been the subject of sustained LGBTQ+ boycott campaigns for over a decade and recently passed 3,000 locations.

People will boycott products that are easy to replace and visible to others. Wearing a competitor’s trainers to signal your values costs you nothing. They won’t boycott products woven into their habits, hidden in their pantries, or requiring genuine sacrifice to replace.

The internet is not your customer base. The people screaming are not necessarily the people buying.

Cornell’s Liaukonytė spent years trying to figure out what actually separates the boycotts that cost brands billions from the ones that dissolve before anyone notices. Social media volume turned out to be nearly useless as a predictor. Anger online tells you almost nothing about what happens at the register. What does tell you something are two much less glamorous questions.

The first: are the people making the most noise actually your customers? A boycott requires someone to stop buying something they were already buying. If the outrage is coming from people who never bought your product and never would, the maximum commercial impact is zero. The boycott pool is capped by your own customer base. The Goya math again — only 7% of American households bought Goya to begin with. The other 93% screaming on Twitter had nothing to withdraw.

The second: how much does switching actually cost? Not money — friction. Effort. Habit disruption. Switching beers is nothing. It’s a different can in the same fridge. Switching away from Amazon means finding another way to handle returns, deliveries, streaming, storage, and whatever your partner ordered last Tuesday. Most people will post about boycotting Amazon. Almost none will do it. The inconvenience is the moat.

Those two questions — who’s actually angry, and how hard is it to leave — produce four very different situations.

Maximum Danger.  Bud Light. Switching beers costs nothing — it’s a different can in the same fridge, same gas station, same sports bar fridge. The people making the most noise were the Tuesday-night, twelve-pack repeat buyers. The ones with something to actually withdraw. They did.

Dead Calm.  Amazon. The outrage is real and has been running continuously since 2012. The commercial damage isn’t, because leaving Amazon requires rearranging your life. The people expressing outrage still need two-day shipping. They post the thread and place the order.

Accelerant.  Nike with Kaepernick. The people threatening to leave were already drifting toward New Balance. Letting them go cost nothing. The customers who stayed didn’t just stay — they leaned in. Controversy this well-aimed doesn’t damage a brand. It sharpens it.

Silent Rot.  The fourth quadrant is the one nobody wants to discuss because it has no clean crisis moment to respond to. The product is too embedded to boycott easily, but the brand has drifted comprehensively away from the people whose money it depends on. Nothing explodes. Everything just quietly gets worse — advertiser relationships strain, talent starts leaving, brand perception erodes in ways that don’t show in monthly revenue reports until one quarter, suddenly, they do. Twitter/X is the live case. No single decision killed it. The platform just kept compounding misalignment with the advertisers funding its operations while being too entrenched to abandon. By the time the numbers reflected it, the rot had been going for years.

Where your brand sits on these two axes isn’t permanent. Customer bases shift. The political context your positioning was built around shifts. A brand safely in Dead Calm in 2019 might be drifting toward Silent Rot by 2025 without anyone in the building noticing — because the revenue reports are still fine, the meetings are still fine, and nobody’s run the diagnostic since the last crisis. That’s the point. Run it before the screenshot arrives.

April 2023. Bud Light sent a personalised can to trans influencer Dylan Mulvaney as part of a relatively minor partnership activation. Conservative media erupted. Kid Rock filmed himself shooting cases of Bud Light with an AR-15. The footage circulated for days. Sales started dropping and didn’t stop. Bud Light lost the number one bestselling beer in America — a position it had held for over twenty years. AB InBev took a $1.4 billion revenue hit.

September 2018. Nike put Colin Kaepernick’s face on a campaign with the line “Believe in something. Even if it means sacrificing everything.” Conservative media erupted. People filmed themselves burning Nike trainers. The hashtag demanded a boycott. Nike’s response was to run the campaign harder. Result: 31% sales increase. Six billion dollars in added market value.

Nike’s core customer skews younger, urban, diverse, progressive. Kaepernick wasn’t a risk for that base — it was a signal that the brand understood them. The people burning Nikes on Twitter were the demo Nike was already losing to New Balance and Adidas. Letting them go cost Nike almost nothing. Signalling to their actual customers cost them nothing either. The maths were clear before anyone lit a match.

Bud Light’s core customer is structurally the opposite. The boycotters weren’t fringe internet agitators. They were the Tuesday-night, twelve-pack-in-the-fridge, repeat-purchase base. Maximum danger, exactly as the framework predicts.

Then Bud Light managed to make it worse by doing precisely what the tense meeting always produces: they found the middle. Didn’t defend the partnership. Didn’t apologise for it either. Just went quiet and hoped it would pass. The result was the worst possible outcome — they lost the conservatives who felt betrayed and the progressives who felt abandoned. Both sides saw cowardice. Neither side saw a brand worth defending.

But even then… Kid Rock coudn’t even resist a Bud Light frothy.

The framework is not an argument that conviction always wins. It’s a diagnostic for knowing when conviction is warranted and when it’s denial wearing conviction’s clothes.

Starbucks, 2018. Two Black men were arrested in a Philadelphia store for waiting for a friend without ordering. The backlash was immediate, justified, and aimed squarely at Starbucks’ stated identity — the progressive, inclusive “third place” that was central to how the brand had positioned itself for two decades. This wasn’t a brand being targeted by people who had never been its customers. This was Starbucks’ own customer base confronting a failure to live up to what the brand had explicitly promised them.

CEO Kevin Johnson issued an apology that was direct and unambiguous — not “we’re sorry you feel that way,” not a statement about the “complexity” of the situation. An acknowledgment that what happened was wrong. Then they closed 8,000 stores for racial bias training, met with community leaders, and made structural changes that were visible and specific. Revenue climbed from $29.3 billion to $36.2 billion between 2020 and 2024.

Compare United Airlines in 2017. A passenger was forcibly dragged off a plane. The initial response was cold and legalistic — defending the airline’s right to remove passengers per its own policy. United only apologised after the video had already been viewed tens of millions of times and the story had permanently embedded itself in how people thought about the brand. The delayed, defensive posture turned a recoverable incident into a defining one.

YouGov’s research on what actually ends boycotts found something brands consistently get wrong: the most persuasive response is showing evidence of what changed, not expressing contrition for what happened. Consumers don’t need remorse. They need proof. Starbucks understood this. United’s first instinct produced the opposite — a statement defending their actions, which is the one response guaranteed to confirm every negative impression already forming.

Run both cases through the framework. Starbucks’ core customers expected progressive values. The Philadelphia incident was a deviation from stated identity — the kind of failure that is forgivable when addressed with evidence of genuine change. United’s core customers expected to be treated with basic dignity. The incident confirmed exactly what they feared about flying, and a late apology for enforcing the policy that caused it changed nothing about what they’d seen. Same crisis structure. Different positioning contexts. Different recovery trajectories.

The brands that come through controversies without being diminished by them aren’t the ones with better crisis communications teams. They’re the ones that built positions so clear and so committed that most controversies can’t find purchase — because the people with the loudest grievances were never their customers and both parties knew it.

The risk most brands refuse to discuss honestly in those tense meetings is not the risk of taking a position. It’s the risk of taking the wrong one — wrong meaning misaligned with the customers whose actual purchase behavior funds the company. Bud Light’s $1.4 billion lesson is right there in the framework. The question the meeting should be asking isn’t whether to respond. It’s whether the brand’s position is genuinely coherent with who it is actually selling to.

Your crisis communication strategy is your positioning strategy under pressure. The brands that come through intact aren’t the ones with better PR reflexes. They’re the ones who already knew, before the screenshot arrived, what they were willing to lose.

If the answer is clear, the decision usually is too. The outrage fades in six weeks. The absence of an answer compounds for years.