Launch Stunts: What Works, What Backfires
Tactics · 10 min read ·
A cautious guide to launch stunts: when a bold gimmick earns attention, how to test it for legal and ethical risk, and the common ways stunts go wrong.
A launch stunt is any attention-seeking act that goes beyond ordinary announcements: a giant billboard, a public challenge, a mock funeral for an old way of working, a live build streamed to strangers. When it works, it is the thing people remember. When it does not, it is the thing people screenshot.
This is a cautious guide. It describes the kinds of stunt that tend to work, the tests that tell you whether yours is safe and the failure modes to avoid. It is deliberately conservative, because the downside of a bad stunt is larger than the upside of a good one.
What a stunt is
A publicity stunt is a planned event designed to attract the public's attention to a person, product or cause. Guerrilla marketing, a related idea, uses unconventional, low-cost tactics to promote a product, often in public places. Both rely on surprise and shareability.
The distinguishing feature is that the stunt itself becomes the story. People talk about it, photograph it and write about it, which gives the product exposure it could not buy through ordinary channels. That also means the stunt is judged on its own merits. A clever idea attached to a weak product will get attention and then disappointment.
Four tests before you commit
Run every idea through four questions. If it fails one, change it or drop it.
1. Is it true?
Does the stunt rely on any statement or implication that is false? Fake scarcity ("only ten licences left" when there is no limit), staged testimonials, invented news and fake outrage all fail. Advertising codes in the UK and guidance in the US treat misleading claims as breaches, whether they appear in a banner or a prank.
2. Is it harmless?
Could anyone be hurt, physically, financially, emotionally or legally? Public space, other people's property, other people's brands and anything that could cause alarm all raise the stakes. A stunt that makes passers-by think there is an emergency is not a stunt; it is a problem.
3. Is it connected to the product?
A stunt should make people think of the product, not just of the stunt. If the audience remembers the spectacle but not what you do, you spent attention and got nothing. Good stunts demonstrate the product's promise: a speed claim shown by racing something, a simplicity claim shown by doing the job in public with one hand.
4. Could you explain it to a journalist, a lawyer and your customers?
If you would be embarrassed to explain exactly what you did and why to any of the three, do not do it.
Stunts that tend to work
Without naming real campaigns, here are the shapes that earn attention honestly.
The live demonstration. Do the impossible-sounding thing in public. Build a site in ten minutes. Process a pile of documents while people watch. It is risky because it can fail, and that is exactly why it is believable when it works.
The public challenge. "Beat our time and win." A fair, clearly worded challenge invites participation. Publish the rules, the judging method and the result, and keep your promise to the winners.
The giveaway tied to the product. A real, clearly described offer, with rules, an end date and an honest account of how winners are chosen. Observe the legal requirements for promotions where you operate.
The open build. Show the work, including mistakes, in public. Transparency itself is the stunt.
The data drop. Publish something genuinely interesting from your own data, such as an anonymised and aggregated finding, with the method explained. Newsworthy and defensible.
The honest provocation. A billboard-sized statement of an opinion you can back up, such as the contrarian stances discussed elsewhere, in a place where people will see it.
Stunts that tend to backfire
Fake news and fake scandals. Staging a controversy that is not real. Once discovered, it overshadows everything.
Fake scarcity and countdowns that reset. Visitors notice. The ones who do not notice may complain to the regulator.
Hijacking a tragedy or a sensitive event. Never. The reputational cost is high and the human cost higher.
Ambush against a competitor. Provocations aimed at a named company can cross into disparagement or unfair comparison.
Stunts that need other people's consent. Using someone's image, work or property without permission invites complaints and removal.
Stunts that depend on deceiving the press. Journalists remember, and they talk.
Stunts that over-promise. A viral moment sends a flood of visitors to a product that cannot yet handle them. Check capacity before you court a spike.
Legal and platform rules
Several areas deserve a check before launch.
- Advertising rules. Claims in a stunt are claims. In the UK, the Advertising Standards Authority's codes require that marketing be truthful and that claims be substantiated. In the US, the Federal Trade Commission's guidance says advertising must not be deceptive.
- Promotions and prize draws. Many places regulate competitions and giveaways; rules about terms, eligibility and fair selection may apply. Seek proper advice for anything involving prizes.
- Public spaces. Permits and permissions may be required for events, projections or signs.
- Intellectual property. Do not use others' trademarks, music or images without the right to do so.
- Platform terms. Social platforms have rules about promotions, automation and impersonation. Breaking them can get your account removed on launch day.
- Data protection. If the stunt collects personal data, your obligations apply from the first sign-up.
This is not legal advice. For anything with real exposure, consult a qualified professional.
Plan for the downside
Write a short plan before the stunt begins.
- What could go wrong? List the five most likely and the one most damaging.
- How will we notice? Who is watching, and on which channels?
- Who decides to stop? One named person with authority to end it.
- What is the first message? A short, plain statement ready to publish if needed.
- How will we fix harm? Who contacts affected people and what do they offer?
A team that has thought through these questions acts calmly under pressure.
If it goes wrong
If a stunt misfires, resist the instinct to defend it. The sequence that works is: stop, acknowledge, explain, fix, learn. Say what happened in plain words, who was affected and what you are doing about it. Avoid blaming the audience for misunderstanding. Follow up later with what you changed. People forgive mistakes handled honestly far more readily than mistakes denied.
Measuring a stunt
Decide beforehand what success means. Visits and sign-ups attributable to the stunt, press mentions with links, quality of the people it brought and, importantly, what happened afterwards. Attention that does not turn into use is a cost. Compare the cost of the stunt, including staff time, to a typical week's results and decide whether to repeat or retire the idea.
A worked example
A small team has built a tool that converts spreadsheets into web forms. Their first idea is a stunt: announce that a famous competitor is "shutting down" and direct visitors to their product. They run it through the four tests. It is not true, so it fails the first test at once. It would also drag a named company into a false story. They drop it.
Their second idea: a public challenge. They post a spreadsheet of a hundred rows of messy event registrations and offer to turn it into a working form, live on video, in under five minutes, and invite anyone to submit their own messy spreadsheet to try to beat the clock. They publish the rules, the way the winner will be chosen and the fact that submitted data will be used only for the demonstration and deleted afterwards.
The tests: it is true, because they really will do it; it is harmless, because participants control what they submit; it is connected to the product, because the stunt is the product's promise; and they could explain it to a journalist, a lawyer or a customer without wincing. On the day the first run takes six minutes, which they say out loud, and the second takes three. The honest stumble makes the stream more watchable, and people share the clip because they can see for themselves what the tool does.
Questions founders ask
Should we hire an agency for a stunt? Only one that will state in writing how they check claims and permissions. The legal responsibility for your marketing stays with you.
What is a reasonable budget? The best stunts are often cheap; ingenuity beats spend. Whatever the amount, spend it on the idea and on contingency, not on volume alone.
How soon after launch should a stunt run? Either as part of launch week, so that attention compounds, or well after, as a second moment. Avoid stacking several on the same day; people cannot follow them all.
How do I know when to stop doing stunts? When the product itself is generating the stories. A mature product is better served by consistent quality than by spectacle.
Summary
A stunt can give a launch a moment ordinary posts cannot. Test every idea for truth, harmlessness, connection to the product and explainability. Favour live demonstrations, fair challenges, honest giveaways and data drops; avoid fake scarcity, manufactured scandal and ambushes. Check advertising, promotion, property and platform rules, plan for the downside and be ready to stop. The stunts that last are the ones you would be proud to describe in full.
Questions and answers
- Are launch stunts worth the risk?
- Sometimes. A stunt that is clearly connected to the product, harmless and honest can earn attention that ordinary posts cannot. One that is deceptive, risky or off-topic usually costs more than it gains.
- What is the first test for a stunt?
- Ask whether it would still be acceptable if everyone involved, including your customers and the press, understood exactly what you did and why.
- Can a stunt break advertising or consumer rules?
- Yes. Fake scarcity, hidden advertising and misleading claims can breach the codes and laws that apply to marketing in many countries.
- What should I do if a stunt goes wrong?
- Stop it, say plainly what happened, fix the harm and explain what you will change. Speed and honesty matter more than polish.