The Zero-Day Marketing Exploits Behind Breakout Startups

Alistair Croll tells Lenny that he has been a product manager all his life, even while running events, and that too many product managers are consumed by the next feature. In his view, many are doomed because they ignore go-to-market strategy and distribution at their peril. The only thing that matters, he argues, is whether a startup has an unfair advantage: a way to capture attention and turn it into profitable demand. That question is often an afterthought or not even a thought.

Croll defines the core move not as doing something evil but as using a system in a way its creators did not intend. He wants founders to remember that distribution matters and that subverting norms to get the system to behave in your favor is important. The book, written with Emily, studies those loopholes and how to get people to pay attention to what you have built, which Croll sees as the hardest part of launching a startup.

Known tactics such as invite lists or Dropbox's two-sided storage bonus work only the first time, Croll says; if they work, they become marketing, and if they become illegal, you stop. The key is not to copy those tactics but to find your own zero-day marketing exploits. He and Emily compared notes across Europe and North America and found tactics that had worked in one market but were untried in another.

Croll explains that the title is tongue-in-cheek because a startup is by definition a disagreement with the status quo, and the status quo was created by those in power. Those in power quickly label a disagreeable startup as evil, even when the startup is trying to do something valuable. He quotes Chris de Burgh's Spanish Train—the Devil still cheats and wins more souls while the Lord is just doing his best—to argue that good causes cannot assume they will win on merit alone. They still have to get the world to behave differently.

The concept originated during Lenny's startup LocalMind at the Year One Labs accelerator in Montreal. The team had validated that strangers would answer geofenced questions—about 95% would answer—but users were not asking questions themselves. Croll suggested fake-asking questions on the platform so people could see what it was like to answer. Lenny objected that he was not evil, and Croll replied that he was not saying be evil; he was saying be just evil enough.

Once LocalMind sent questions from the app to people in a location, the team realized it could collect passive data about what was happening in a bar or restaurant—similar to what Google Maps later did. The fake questions also showed new users what kinds of questions to ask, a common fix for cold-start problems. The story captures the book's boundary: provoke the change required for the product to succeed without crossing into actual harm.

Croll's method starts with system awareness: understanding the system you are in and finding a novel approach within it. Then comes disagreeability—being willing to do something others cannot or will not do. He insists that subversiveness is a skill, not a personality trait, and that he and Emily studied neuroscientists and historical examples to understand it. The three ingredients are system awareness, novelty, and disagreeability.

His favorite system-awareness exercise comes from Stanford professor Tina Seelig, who gave engineering entrepreneurship students $5 in seed capital, five days to plan, and two hours to execute. Most teams bought lottery tickets or set up lemonade stands; some pumped bike tires for donations and made about $200, while another team sold places in line at fancy restaurants and made about $250. The winning team made roughly $650 by selling its three-minute presentation slot to a company that wanted to recruit Stanford graduates. They saw the system they were actually in—a room full of recruiters—rather than the resources they were told they had.

Croll uses Gymshark's Jim Shark birthday campaign—B-list celebrities wishing a fictional person happy birthday around the time Cameo launched—as an example of novelty. He also cites reverse graffiti, such as the British Intelligence Agency spraying a recruiting ad on a sidewalk, and Coinbase's 60-second Super Bowl ad that bounced a QR code around the screen. Coinbase was ranked the worst of 64 Super Bowl ads but got 20 million hits, and its servers fell over. The lesson is to zag when everyone else zigs.

Lenny asks whether such tactics are just turbo boosts that fade, and Croll agrees that many are. The companies that do well use the same thinking to change their value chain. IKEA saw that flat-packing furniture and letting customers assemble it at home would change how furniture was made, shipped, and sold, delegating one stage of creation to the customer. That is not a one-off stunt; it is a fundamental change to the business model.

Croll distinguishes these moves from growth hacks, which he says are product-agnostic—like the pop-up asking why you are leaving a website. He prefers zero-day marketing exploits because they require understanding the specific dynamics of an industry, the difference between a script kiddie and a proper hacker. The point is to find approaches that are fundamental to the business model, not just traffic spikes.

Croll cites researcher Zachary Hambrick's Navy experiment, where sailors monitored four consoles for a score. Halfway through, the scoring changed so one quadrant represented 75% of the points; some sailors noticed and hammered that quadrant, while others conscientiously kept doing what they were told. The Navy has historically recruited conscientious people who follow instructions, but Croll points out the difference between playing the game right and questioning whether you are playing the right game.

Poker is his analogy: the rules tell you which hands win and whose turn it is, but you win by bluffing, subterfuge, and reading tells. Business schools rarely teach that kind of thinking. He tells the story of Deu-la-Deu Martins, who was left to guard a castle with about two weeks of food; she baked all the flour into bread, threw it over the parapets to the besieging army, and said, 'You look hungry. Let us know if you need more.' The army left, and the town's coat of arms is two crossed loaves of bread.

Lenny brings up Rory Sutherland's advice to spend time on the rational things everyone agrees on, then set aside time for the irrational. Croll loves Sutherland's wine-list example: a restaurant can sell wine at almost any price because customers cannot make it that way, and once the list is handed to the most powerful-looking person who says red or white, everyone has consented to order wine. Sutherland, he says, sees the world through system lenses better than almost anyone.

After the subversive mindset, Croll says, you have to scan the market. He and Emily created the Recon Canvas, which splits the market into product, medium, and market. His hot take is that product-market fit is overrated: two companies can sell the same product to the same market, and one wins because of its medium strategy. Marketing textbooks were written when the medium was one-way, one-to-many paid broadcast; today it is any-to-any, and it is valid to build an audience first, then figure out what to sell, then make the product.

The medium includes both the platforms you are in and the norms of those platforms. Burger King is his model: when it launched its mobile app, it offered a free Whopper to anyone who ordered from the app and picked it up—but only if they ordered from a McDonald's parking lot. That drove installs and geolocation. In 2019, Burger King also liked influential people's posts from 10 years earlier and offered no explanation; because of platform mechanics, those old likes were noticed in a way a like on a recent post would not have been. The reveal was that funnel cake fries were coming back, and the stunt earned millions of free impressions.

The canvas has three rows—objective, collective, and subjective—and sections for product (features and messages), medium (platforms and norms), and market (attention and actions), for 18 squares total. For each square, the question is: what is the status quo, and where is there an opportunity to do something unorthodox? Croll says the book contains about 160 case studies, with five or six for each rectangle.

Lenny offers Runway's locked bag as a recent example: influencers received a bag with a countdown lock that would open in three days, creating a forced unboxing moment on Twitter. Croll shares a Lean Analytics launch hack: he and Ben Yoskovitz scheduled about 15 supporters to tweet over seven days, every 12 hours, each with a UTM tag so they could compare who drove clicks versus pre-orders; then they wrote a blog post explaining how they hacked the launch, which proved they knew analytics. The through-line is that understanding platform mechanics and norms is part of go-to-market.

Croll and Emily reviewed 300 to 400 case studies—about 160 made it into the book—and found 11 tactics that keep showing up as patterns. The first is turning a bug into a feature: Salesforce's limited AJAX product became simple against million-dollar systems like Vantive, Clarify, and Remedy, and its slogan was No Software. The second is buyer upgrade: Croll offers the apocryphal example of Mr. Clean's Magic Eraser, which began as melamine foam for aircraft insulation until people used it to clean stains; Hitachi's personal massager found a market through Vibratex on condition that Hitachi not be mentioned; and bridge-inspection drones found a better buyer in insurance companies. The third is access: Getaround's Jessica Scorpio used a VC friend's early Tesla Roadster at CES, MasterClass's founder went to school with Dustin Hoffman's daughter, and Bumble's Whitney Wolfe Herd used her sorority network to get women on the app before telling fraternities the women were there.

Bait and switch works only if the buyer is delighted with the switch or also gets what was promised. Tupperware promoted dinner parties to women who wanted to run their own businesses, then switched them into a multi-level marketing scheme, while Energage used local newspapers' Best Workplaces surveys to get companies to fill out HR surveys and then called them with the data. Combination is another tactic: Kraft could not sell powdered cheese alone, so it paired the cheese with macaroni to create a ready-made dinner, and 1-800-MATTRESS added old-mattress removal because New Yorkers did not want a new mattress so much as a replaced one. Arbitrage means knowing something others do not: bankers paid an operator on the Chappe optical telegraph to introduce an error in Paris so they could trade on the news before it arrived, and Farmville grew because Mark Zuckerberg told Mark Pincus that apps would be allowed to post to users' timelines.

Aggregation takes data from many places and modifies it. Busbud's LP Maurice collected bus schedules from around the world, automated millions of pages of structured JSON data, became the default search destination for bus routes, and then used the traffic he sent to bus companies as leverage to sell tickets. Get Satisfaction, by contrast, crawled complaints about companies like Sony and then threatened to tell customers the company did not care unless it bought from them—an example Croll considers too evil. Reframing draws a new grid rather than competing on the old one: Tom's of Maine positioned unfluoridated, natural toothpaste as orthogonal to Colgate's sex appeal and Crest's clinical claims, and Tesla raced electric cars against Bugattis to make performance, not just range and sustainability, the relevant dimension for a segment.

Regulation can be a strategy: Austria's opt-out organ donation system yields about 99.6% donors, while Germany's opt-in system yields about 12%, so changing the default can do more than any growth hack. Parklets emerged when urban activists built temporary parks in parking spots because the law did not say a car had to be there; one Englishman built a park on a car, then bought a historical van exempt from smog checks when the town objected. Sliding the window changes what is considered acceptable: Bodyform was the first to use red liquid in ads for feminine hygiene products after it had literally been illegal, and the campaign made women feel someone had finally said the quiet part out loud. Croll also lists misappropriation—Netflix using the US Postal Service as a high-latency broadband network by mailing DVDs—as another recurring pattern.

Croll says operationalizing this is somewhat magic and takes time, but it is a muscle. He advises spending time learning to be disagreeable and understanding the system, then temporarily letting yourself think like a super villain by ignoring guardrails, embracing absurdity, and not pulling punches on wording. The Blair Witch Project team listed actors as missing, underfed them, played scary noises outside their tent, and gave them conflicting information, generating hyperbolic attention. He also points to TRIZ, construal level theory, Wes Kao's spiky point of view, pre-mortems, and counterfactuals—asking whether the opposite is true, whether this would be better a year ago or a year from now, and what a smart coach would conclude from the same information.

Croll has already applied the tactics to the book launch. Early on, he and Emily ran a survey for pricing and customer research and offered a free consulting session as a prize, but realized respondents were disincentivized to share it. So they created Team Orange and Team Black and said the winner would be chosen from the team with the most submissions, prompting people to recruit for their team. The book also uses QR codes next to case studies to link to videos, resources, and a list of weak signals for finding exploits; traditional publishers dislike QR codes and authors retaining intellectual property, so they chose an unusual publishing path.

The book devotes a chapter to Don't Actually Be Evil. Croll says not to abuse people, assume consent, act without approval, lie, use dark patterns, break the law or its intent, or ruin your reputation. He cites Afterlife, which crawled obituaries and led funeral homes to send flowers years after funerals; its operators were charged in criminal court. He also cites the Chrysler PT Cruiser, which was technically a truck and helped Chrysler's truck line meet fuel-efficiency standards so it could keep selling bad trucks—legal, but too evil for the book.

In the lightning round, Croll recommends Dan Davies's The Unaccountability Machine, Jonathan Haidt's The Righteous Mind, and Dan and Chip Heath's Made to Stick as the best marketing textbook. His recent favorite movie is David Fincher's The Killer, which he calls the embodiment of self-awareness, and his favorite recent product is a $100 folding second screen for his MacBook. His company is called Solve for Interesting, and his motto is that it is amazing what can get done when nobody cares who gets credit. He also recalls Lenny moving from San Diego to Montreal and jumping into winter life, including tobogganing and learning the words toboggan and toque.