The Man Who Named Growth Hacking Says You’re Doing It Wrong

Sean Ellis has a confession that sounds strange coming from the man who invented the phrase every startup marketer has chased for a decade. “Most companies fail with growth hacking,” he says. “It’s unfortunate, but it’s the reality.” He is not being coy. Ellis coined the term roughly twelve or thirteen years ago, watched it circle the globe, and has spent the years since watching most of the people using it get it wrong.

The wrongness, he insists, is not a shortage of cleverness. “I don’t think it’s because they have a hard time coming up with a hypothesis or a hard time coming up with the right test idea,” he says. Modern digital marketers are smart. They test. They optimize. The problem is that everyone else got smart too — and when everyone bids on the same auction, Facebook and Google simply raise the price. “Anytime someone gets good at these things, the prices keep going up,” Ellis says. “It’s harder and harder to find a profitable way to get people to the website.”

So marketers get frustrated. They give up on running experiments inside the product and retreat to the channels they can control. “Then it’s not growth hacking anymore,” Ellis says flatly. “They’re just doing marketing.”

01The accidental heretic

Ellis did not set out to name a movement. He started in sales. When an early employer handed him the marketing job, he protested that he had never studied it, then enrolled at New York University to fix the gap. He earned an A. And then, back at his desk, he was terrible. “The class kind of broke me,” he says. “I became too academic.” He had to consciously purge the textbook and return to the only thing that had ever worked for him: getting people to actually use a product.

That instinct hardened after he moved to Silicon Valley in 2007, where venture capitalists kept introducing him to founders who couldn’t grow. The founders always opened with the same request. “The first thing they would say is, ‘We really need your help to build awareness,’” Ellis recalls. He never bought it. “I’ve never focused on building awareness. I want them to actually use the product — sign up, use it, get value, pay me money, and then I take that money to get more customers.”

That was the moment he realized his approach needed a different name. Not marketing. Something scientific. He defines growth hacking as “a scientific approach to figuring out how to grow the business” — testing and analyzing across every lever from acquisition to activation to retention to monetization. The reputation came from results: he joined Dropbox when it had fewer than ten employees and helped make it the fastest SaaS company ever to reach a billion dollars in revenue. He later ran interim growth at Eventbrite and worked with Lookout. All three crossed a billion in value.

02The lever nobody watches

Ask Ellis where he starts, and he does not say advertising. He says activation — the first real experience a user has with the product. “Activation is like the most important one that I would generally focus on,” he says. “It’s the first one I focus on when I work with a company.” The logic is a chain: the most powerful thing that brings people back is a great first experience, so a strong activation step is the biggest driver of the retention and engagement that everyone frets about later.

This is also where he dismantles the caricature of the growth hacker as a con artist. “The best growth hacks are not things that trick people,” he says. “They’re finding where people get confused in the product today and finding a way to make it easier and simpler.” When you build something, he notes, you’re often just guessing at what will work. Sometimes the guess is good. Often it takes thousands of real users bumping into the same confusion before you understand what to fix.

03Why the smart ones still fail

If the ideas are fine and the tactics are known, why do most companies fail? Ellis’s answer is structural, almost boring, and that is exactly why he thinks people miss it. “Growth is cross-functional,” he says. It requires product, marketing, engineering, design and data people to work together. “Most companies aren’t organized that way. As they get bigger, companies essentially have more and more specialist teams that become separated. That’s what causes growth hacking not to work in most businesses.”

The failure has a predictable shape. It turns out the product team has more growth leverage than the marketing team, because the biggest opportunities to improve and retain a customer live inside the product. But a marketer who wants to run a product experiment has to spend half their energy just convincing the product team to let them. “They get frustrated, so then they stop trying to run experiments in the product,” Ellis says — and they drift back to Facebook and Google, back to plain marketing. The resistance is usually fear dressed as principle: brand designers worried the testing will make things ugly, product people worried the “growth hackers” will make the product awful. “That’s a perception thing,” he says.

04Take the shot — then take fifty more

Once the teams cooperate, the work becomes a loop: analyze who loves the product and why, generate ideas, prioritize them, run the test, read the results, repeat. Ellis returns again and again to one word — velocity. “Nobody knew consistently what was going to be the best,” he says of his early wins. “The only way to figure out what was the best was just to test a lot of things.” He cites a line he attributes to Amazon’s founder: the company’s success is a function of how many experiments it runs per day, per month, per year.

His favorite framing comes from the ice rink. Wayne Gretzky said you miss every shot you don’t take, and Ellis extends it to soccer, to any sport, to testing itself. “If you only take one shot, maybe you make it. But if you take fifty shots in a game, you’re much more likely to win.” The mistake teams make is falling in love with a single perfect test. Better, he argues, to run three or five different tests every single week — because you genuinely cannot predict which one becomes the outlier that changes the trajectory of the business.

05The one question that decides everything

There is a prerequisite to all of it, and Ellis is unsentimental about it. “You can’t grow something that people don’t really like when they try it,” he says. In startup language: product-market fit. He waves off the modesty reflex. “A lot of times I’ll tell people I got lucky, and they’ll say, ‘Oh, you’re just being humble.’ No — if you give me something to grow and nobody likes that product, I will fail every time.”

To measure it, he built a question now known as the Sean Ellis Test. He surveys people who have actually used a product more than once, ideally recently, and asks: how would you feel if you could no longer use this? If only 5% say they’d be very disappointed, he walks away — there’s nothing to grow. But across hundreds of companies he found a threshold. “About 40% of the users said they would be very disappointed without the product,” he says, “and those companies were generally successful.” Forty percent became his leading indicator.

The harder truth lives in the retention curve. Start with 100 new users. If the line slides toward zero, there is no product-market fit — you’re only replacing the people who leave. But if it drops to 70, then 60, then 50 and plateaus — running flat and parallel to the x-axis — that flat line is the signal. Where it flattens depends on the business: the meditation app Calm plateaus near 5% because habits are hard; Instagram flattens closer to 50 or 60%. “It’s less important where it plateaus,” Ellis says. “But if it always goes to zero, eventually you’re not growing — you’re just replacing.”

This is also why Ellis is strategic about which companies he joins. “The best company for me to work on is one that has all the signs of product-market fit but no growth yet,” he says. The reason is coldly rational: the moment a company starts growing, investors declare it valuable and pour in money, and the stock gets expensive. “If I see signs of product-market fit and it doesn’t have the growth yet, that stock is still very cheap.”

06The quieter payoff: humility

The most surprising benefit Ellis names has nothing to do with metrics. “One of the biggest impacts of growth hacking on the culture of a business is that everyone becomes less arrogant,” he says. When people who thought they had the answers watch tests prove them wrong a good fraction of the time, they stop insisting and start wondering. “They start being curious about what is the right answer,” he says. The arguments about my opinion versus yours give way to a shared question: how do we actually uncover the truth?

It is a fitting end for a philosophy that began when a marketer with a fresh A in marketing decided the textbook was the problem. The word he coined went on to be misused by a generation that mistook it for a bag of tricks. Ellis keeps pointing them back to the boring, powerful center of it: love the product first, test everything, take more shots, and let the results — not the loudest voice in the room — decide.