Which six cases show how stunts work?
Six cases show how stunts work across three decades: Taco Bell in 1996, Red Bull Stratos in 2012, the ALS Ice Bucket Challenge in 2014, the Cards Against Humanity Holiday Hole in 2016, Liquid Death from 2019 and Stanley in 2023. Each has a published, checkable result.
| Case | Year | What happened | Reported result |
|---|
| Taco Bell Buys the Liberty Bell | 1996 | Full-page ads in six newspapers said Taco Bell had bought and renamed the Liberty Bell | Thousands of calls to Taco Bell and the National Park Service, and a reported sales rise of about $1 million over two days |
| Red Bull Stratos | 2012 | Felix Baumgartner's freefall from 38,969 meters above New Mexico, streamed live | More than 8 million concurrent YouTube viewers, the previous record was about half a million |
| ALS Ice Bucket Challenge | 2014 | A participatory challenge in which people poured ice water over themselves or donated | $115 million donated to the ALS Association over eight weeks |
| Cards Against Humanity Holiday Hole | 2016 | The company livestreamed a hole being dug on Black Friday for as long as donations lasted | More than $100,000 raised, and more than 50 hours of digging |
| Liquid Death | 2019 onward | A canned-water brand built on entertainment-style marketing | A $1.4 billion valuation in March 2024 and $263 million in retail sales in 2023 |
| Stanley Quencher | 2020 to 2023 | Influencer seeding, collaborations and limited drops around a tumbler the company had stopped marketing | Revenue up from $73 million in 2019 to a projected $750 million in 2023 |
Taco Bell Buys the Liberty Bell (1996)
On April 1, 1996, Taco Bell ran full-page newspaper ads announcing that it had bought the Liberty Bell and renamed it the "Taco Liberty Bell." The Philadelphia Inquirer's retrospective describes switchboards lighting up at the paper, at the National Park Service and at Taco Bell's headquarters, and reports that sales rose by about $1 million over the next two days. The company's agency, Paine PR, had worried that someone would reject the ad or the press release before it ran.
The Park Service held a news conference to deny the sale, and its director at the time, Roger Kennedy, called the ad "as false as it is cheesy," according to the Museum of Hoaxes archive. Taco Bell confirmed the joke at midday. The stunt worked because the calendar gave the press permission to play along and because the brand resolved the joke the same day.
Red Bull Stratos (2012)
On October 14, 2012, Felix Baumgartner jumped from a balloon capsule at 38,969 meters above New Mexico and reached 1,357.6 kilometers per hour, becoming the first person to break the sound barrier in freefall, per Guinness World Records. Guinness describes the jump as the conclusion of a five-year, $20 million Red Bull Stratos project. Other reports cite higher figures, so the cost remains an estimate.
The live stream carried the story. The Wharton School's knowledge site reports that YouTube's director of project management said the stream accounted for 8% of all internet traffic at its peak, and that the audience grew from 500,000 at liftoff to more than 8 million as Baumgartner stepped out. Red Bull streamed the event on its own channel, which is the owned-IP format in practice: the brand produced the event, owned the imagery and received the credit.
ALS Ice Bucket Challenge (2014)
The ALS Ice Bucket Challenge was participatory rather than brand-led. Participants poured ice water over their heads or donated, then nominated others. According to the ALS Association, $115 million was donated during an eight-week period in 2014, and $77 million of that went to research.
The case shows that a stunt can be designed as a mechanic rather than an event. A simple action, a clear signal and a nomination structure let the audience carry the campaign, and the organization supplied the cause and the donation path.
Cards Against Humanity Holiday Hole (2016)
On Black Friday, November 25, 2016, Cards Against Humanity livestreamed a hole being dug and said it would keep digging as long as people paid. NPR reported that donors gave more than $100,000 for a hole with no stated purpose. Tubefilter's report said the dig ran for more than 50 hours and that the video drew more than 3.6 million views in three days.
The stunt worked because it mocked the shopping day it was timed to, fit the company's irreverent positioning and cost little to produce. The company had run a Black Friday gag every year, so the audience expected one.
Liquid Death (2019 onward)
Liquid Death, founded by Mike Cessario, sells canned water with heavy-metal-style branding and entertainment-style marketing. In March 2024 the company announced a $67 million round at a $1.4 billion valuation and reported $263 million in retail sales for 2023, according to Retail Dive. EPR's profile covers the company's communications in Liquid Death public relations.
The case is a brand-level stunt, not a single event. Each product launch and partnership is designed to be a story, and the positioning keeps the stunts coherent.
Stanley Quencher (2020 to 2023)
The Stanley Quencher launched in 2016, and by 2019 the company had stopped restocking and marketing it. After Terence Reilly, who had worked at Crocs, became Stanley's president in 2020, the company leaned on influencers, collaborations and limited-edition drops. CNBC reported that revenue rose from $73 million in 2019 to a projected $750 million in 2023, and that Reilly described the approach as collaboration culture and drop culture.
Stanley is less a planned stunt than an example of a brand amplifying attention it did not originate. The lesson is operational speed: the company turned social attention into products, drops and collaborations quickly enough to keep the story going.
Five structural formats cover most PR stunts that work at scale: the April Fools' stunt, the founder as stuntman, the owned-IP stunt, the apology stunt and the participatory stunt. Each has a canonical case, a pattern that makes it work and a trap that makes it fail.
| Format | Canonical case | Why it works | The trap |
|---|
| April Fools' stunt | Taco Bell and the Liberty Bell, 1996 | The calendar gives the press permission to play along, and the brand confirms the joke the same day | It works only on April 1, and audiences now expect the joke |
| Founder as stuntman | Richard Branson serving as an AirAsia flight attendant in 2013 | The founder takes visible personal risk, and the press has a face to photograph | It stops working when the founder steps back, and it suits only founders the press already wants to cover |
| Owned-IP stunt | Red Bull Stratos, streamed on Red Bull's own channel | The brand owns the production and the imagery, so the press credits one name | It is expensive, and half-funded versions read as failed sponsorships |
| Apology stunt | The Domino's Pizza Turnaround, December 2009 | Self-deprecation is paired with a fix the audience can try | It suits product or operational failures the brand is clearly at fault for, and draws backlash when used for moral or political missteps |
| Participatory stunt | The ALS Ice Bucket Challenge, 2014 | The audience does the marketing, and the organization supplies the mechanic | The organization has to give up control of the script, or participation stalls |
The founder format in practice
Branson's 2013 stint followed a bet with AirAsia's chief executive Tony Fernandes over which of their 2010 Formula One teams would finish ahead. CNN reported that Branson, in uniform, served drinks and meals on an AirAsia X flight from Perth to Kuala Lumpur, and that the flight was a charity event. The risk was reputational and personal, which is why the press covered it.
The apology format in practice
Domino's put its own executives on camera in December 2009 reading customer criticism, including that the crust tasted like cardboard, and announced a new recipe. The company reported a 14.3% rise in domestic same-store sales in the first quarter of 2010, according to USA Today as reported by PMQ. A Michigan business outlet reported that Patrick Doyle called the move a gamble. The format worked because the fault was real, the voice was candid and a product fix backed the message.
Choosing a format a brand has earned
The practical question is which format a brand has earned the right to use. The April Fools' and apology formats are open to most brands, because the first depends on a free calendar moment and the second on a real failure the brand can fix. The founder format needs a founder the press already wants to photograph. The owned-IP format needs production capability and sustained budget. The participatory format needs a mechanic simple enough for strangers to copy.
Which stunts backfired?
Stunts and stunt-like campaigns backfire when they misjudge the audience, the physics or the law, and the failure then attaches to the brand. Three documented cases show the range: a tone-deaf ad, a consumer boycott and a staged hoax.
Pepsi and Kendall Jenner (2017)
Pepsi released a commercial on April 4, 2017 in which Kendall Jenner joined a crowd of protesters and handed a police officer a can of Pepsi. Critics said it trivialized social justice protests and echoed a photograph from a Black Lives Matter demonstration. On April 5, the company said it was pulling the ad, and its statement, reported by the Associated Press, said it had been trying to project a message of unity, that it had missed the mark and that it apologized. Pepsi also apologized to Jenner. The case is covered in full in EPR's Pepsi case study.
The company's quick reversal limited the damage, and the case remains a standard reference for misjudging a cultural moment. The ad was an advertisement rather than a stunt, but it shows what happens when a brand borrows the imagery of a movement without the standing to do so.
Bud Light and an influencer partnership (2023)
An April 2023 Instagram partnership between Bud Light and influencer Dylan Mulvaney drew a consumer boycott. According to NBC Los Angeles, citing NielsenIQ data, Modelo overtook Bud Light as the top-selling beer in the United States in May 2023, with Bud Light's sales down 24.6% year over year in the four weeks ended June 3. The Associated Press noted that Bud Light had been the top seller since 2001. Anheuser-Busch's chief executive said in April that the company never intended to be part of a discussion that divides people.
The partnership was not a stunt in the strict sense. It belongs here because it shows how a single social post can produce sustained commercial consequences when the audience reads it differently from the brand's intent. EPR's analysis is in lessons from Bud Light.
Balloon Boy (2009)
On October 15, 2009, Richard and Mayumi Heene told authorities their six-year-old son Falcon had floated away in a homemade helium balloon. Rescue operations began and the event ran live on television. The boy was found at home, and investigators concluded the incident was staged. According to a History.com report, Richard Heene pleaded guilty to a felony charge of attempting to influence a public official and was sentenced to 90 days in jail, and he was ordered to pay $36,000 in restitution. NPR reported that 60 of those days were on work release.
The case is the clearest line in the field: a stunt that deceives authorities or diverts emergency services produces criminal liability, not coverage.
What are the mechanics of a working stunt?
A working stunt has six structural elements: it is the story, it has a visual asset, it carries surprise or risk, it fits the brand, it can be defended if questioned and it has a rehearsed crisis response. The six cases above show each element at work.
| Element | What it means | Where it shows up |
|---|
| The stunt is the story | A journalist can write the headline in five words and the lede in two sentences | "Taco Bell Buys the Liberty Bell" and "man jumps from the edge of space" |
| It has a visual asset | Photographs and video produce coverage that text alone cannot | The Stratos live stream and the Holiday Hole livestream |
| It carries surprise or risk | The audience does not expect it, and the brand risks something real | Baumgartner's jump and Branson's flight |
| It fits the brand | The stunt extends the brand's existing position | Red Bull and extreme sport, Liquid Death and irreverence |
| It is defensible | The brand has an answer when the press asks why | A charity tie-in or a stated purpose, as in the AirAsia flight |
| It has a crisis response | A plan exists for what happens if it goes wrong | Pepsi's same-day reversal in 2017 |
The sixth element is the one most often missing. Any stunt can fail, and a brand that has not agreed in advance who speaks, what is said and when the stunt is stopped will improvise under pressure. EPR's crisis PR pillar covers how to build that plan.
How do you decide whether to run a stunt?
A brand should run a stunt only when it can answer yes to all six mechanics questions and clear a legal and safety check. If any answer is no, the idea is probably an expensive advertisement or a risk the brand has not priced.
| Question | A passing answer | A failing answer |
|---|
| Can a journalist write the headline in five words? | Yes, without explanation | It needs a paragraph or a campaign brief |
| Is there a visual that stands alone? | A photo or video that tells the story | The story lives only in a press release |
| What is at risk? | Something real, named in advance and acceptable to lose | Nothing, or something the brand has not thought about |
| Does it extend the brand's position? | The stunt makes sense coming from this brand | The stunt could belong to any brand |
| What does the brand say when asked why? | One clear, true sentence | Hedging or an invented rationale |
| What happens if it goes wrong? | A written plan with named decision makers | The team will work it out on the day |
| Is it lawful and safe? | Permits, insurance and safety reviews are in place, and nothing deceives authorities | The stunt relies on a false report or an uncontrolled risk |
Run the test before the budget is set. A stunt that fails the first two questions will not earn coverage regardless of spend, and one that fails the last two can cost far more than it earns.
How should a team measure a publicity stunt?
A team should measure a stunt against a baseline set before launch, using three measures: coverage earned, search interest and business effect. Without a baseline, any number reported afterward is a claim, not a result.
Coverage earned means the count and quality of unpaid mentions in outlets the audience reads, recorded with links. Search interest means the change in branded and event-related searches during and after the stunt, which a search console and a trends tool both show. Business effect means sales, sign-ups or donations over the same period, compared with the same days in the prior period, as the reported figures for Taco Bell, the ALS Association and Domino's do.
Be careful with equivalent-advertising-value claims. A figure for what the coverage would have cost as an ad is an estimate made by the party reporting it, and it does not show that any sale happened. Where a source gives a business result, such as the $115 million raised by the ALS Association, use that figure and attribute it.
Which campaigns have full EPR case studies?
EPR publishes full case studies on individual campaigns, each with a results scorecard that separates documented figures from claims. Three are live now, and others join this list as they publish.
KFC is a crisis communications success rather than a planned stunt, and EPR lists it here because it shows how candor and humor can work in an apology. Barbie is a coordinated launch campaign, and it is a useful contrast with the single-event stunts above.
What does the history of the publicity stunt show?
The history of the publicity stunt shows the same mechanism moving to each new distribution surface: the newspaper, television, social media and short video. The tools changed and the logic did not.
P.T. Barnum ran stunts as a business model in the nineteenth century, planting stories and manufacturing controversy to draw crowds. Edward Bernays staged a women's smoking demonstration at the 1929 New York Easter Parade for the American Tobacco Company, and the episode is taught both as a stunt and as an ethical warning. EPR's profile of Edward Bernays covers it.
Television and the daily newspaper carried the stunts of the twentieth century, including Taco Bell's 1996 ad. Social media shortened the cycle from days to hours, as the ALS Ice Bucket Challenge showed in 2014. Short video now rewards stunts designed for a phone screen, as Liquid Death's marketing and the Stanley story show.
Where do stunts appear in other EPR case studies?
Stunts and stunt-like tactics appear across EPR's case studies in outdoor advertising, guerrilla marketing, games, sponsorship and festivals. Each shows the mechanics above in a different setting.
The greatest brand stunts executed outdoors covers activations in public space. The guerrilla marketing playbook defines the lower-cost end of the discipline. The I Love Bees case shows an immersive campaign built for a game launch. The Airwallex and McLaren takeover shows a sponsorship activation. Festivals offer the opposite lessons: the Coachella case runs on other people's publicity, the Glastonbury case needs very little, and the Burning Man case rejects brand activation altogether. For the wider discipline, see event marketing and public relations.
Conclusion
A publicity stunt works when one clear event, backed by a real risk and a position the brand has earned, becomes a story a journalist would report unpaid. Every verified case here, from Taco Bell to Stanley, passed that test, and the failures failed it. Run the six questions before the budget is set, plan for the stunt that goes wrong, and set a baseline so the result can be measured.
