Kevin Ryan on Tim Ferriss: 10-Year Trends and the Second Mouse Rule

Kevin Ryan, founder of AlleyCorp and co-founder of MongoDB, Business Insider and Gilt, tells Tim Ferriss how he spots 10-year trends and builds companies on them. In this 76-minute episode of The Tim Ferriss Show, the man often called the godfather of New York tech walks through the wins (DoubleClick, MongoDB, Valar Atomics, Transcend Therapeutics), the ones that got away (YouTube), and the ones that fell apart (Gilt Groupe, a Shopify for healthcare), and explains why his venture fund stays small on purpose.

TLDW

Ryan bets only on trends he thinks will run for ten years, because that is how long it takes to build an important company, and then asks what has to exist if the trend holds. He starts companies when an idea survives a two-week “business crush”, puts product ahead of financial models, fundraising and exits, and expands one narrow step at a time. His current bets are nuclear energy, psychedelic medicine, robotics, AI and healthcare, and he says he has not touched e-commerce since Gilt because that problem is solved. He closes on vacation, ping pong, income inequality and why the United States should want far more skilled immigrants.

Thoughts

The second mouse gets the cheese, and Transcend Therapeutics is the cleanest example of that rule in the whole conversation. Ryan started it in 2021, after the first wave of psychedelic companies had already shown where the traps were. Ferriss names two of them: a treatment that lasts four to six hours needs extra nurses and rooms, which puts the cost out of reach, and bundling the drug with psychotherapy gave FDA advisers something they could not standardize. Methylone was picked because it avoids the first trap by chemistry. It is shorter and gentler than MDMA, and in Transcend’s trials it could be given once a week. That is a choice made on how well a drug fits a clinic’s schedule, not on which molecule is the most impressive. A ten-year trend is long enough that you can let someone else find the trap first. The catch is that being second only pays if you change something that matters, and Ryan’s own Gilt, a copy of a French company, shows what happens when you do not.

“Big funds focus on the 2%. We focus on the 20%” is the most useful sentence here for anyone trying to judge a venture fund. Ryan says none of AlleyCorp’s three funds has come in under a 50% IRR, so he could raise several times the $335 million he just closed, and he declines. He is also the largest investor in his own fund. Put those together and you have a manager whose income depends on the companies working and not on the size of the pile. The same logic explains a policy most venture investors would call heresy: four or five years in, AlleyCorp now plans to sell about a third of a winning position. The returns are his own figures and nobody on the show checks them, but the structure is worth copying regardless of the numbers.

Gilt is the proof that a correct trend is not a business. Online luxury discounting grew exactly as Ryan expected, to $500 million in revenue by year four. Then the brands built their own websites and discounted their own stock, the department stores caught up, and Gilt was one buyer among many. He could take 1,000 of the 20,000 items Theory had left at the end of a season, which was not enough to matter to Theory. The part worth studying is what he did next. He told his board to sell a company once valued at a billion dollars, hoped for $400 million, took $250 million, and three years later the buyer offered it back for $5 million. A trend tells you a market will exist. It says nothing about whether you get to keep your place in it, and knowing when to sell a falling knife is a skill founders rarely get credit for.

Ryan’s filter for robotics is better than most of what gets written about humanoids. He expects 100 to 10,000 times more robots in ten years, and he thinks the money is in machines that do one thing in one place, a textile factory or a car plant, because a restaurant worker who chops vegetables, hauls garbage upstairs and runs to the deli is still far beyond a robot. His second test is unfilled jobs. He cites 17,000 open massage therapist positions in the United States two years ago, which is why he backed a robot massage company. Automating work nobody is available to do means there is no incumbent to fight and no worker to displace, and the customer was already going without. That is a far easier sale than replacing someone.

The honest weakness in the “business crush” method shows up when Ryan puts two slow starts side by side. MongoDB had no revenue for three and a half years and only about $40 million after eight, and he says it is now worth roughly $30 billion. His Shopify for healthcare sites absorbed $6 million and was shut down, and he admits he still does not know why. From the inside the two looked alike for years. The one difference he gives is that MongoDB’s usage kept growing fast while the revenue did not. If you are funding something on a ten-year view, usage is the number that tells you whether to keep waiting, and revenue will mislead you in both directions.

Key Takeaways

Bet on trends that will last ten years. An important company cannot be built in two or three, and anything that is hyped today or started five years ago is already too late.
Ryan’s whiteboard exercise is to list 20 trends, then ask what will exist in ten years that does not exist today, and who will supply the winners (the second-order bet).
In 2003 online video cost about $10 per thousand views to serve and earned about $1 in advertising. His team saw the two lines would cross around 2005 and still did not start YouTube.
Gilt Groupe reached $175 million in revenue in year two and $500 million in year four, then sold for $250 million after brands began discounting on their own websites.
DoubleClick opened offices in 25 countries in three years while competitors were in six. Global advertisers such as Microsoft and Procter & Gamble signed up for the footprint, and smaller clients followed them.
The original AlleyCorp model: Ryan and Dwight Merriman each put in about $500,000, built and launched the product for roughly a year, then raised venture money. Three of the first six were Gilt, Business Insider and MongoDB.
Start narrow. Business Insider began with three people covering New York tech and grew one vertical at a time to 600 journalists. Gilt began with one women’s clothing sale a week.
He does not build a financial model, plan the fundraise or think about exits at the start. If the product is one everyone wants to use, the money and the hires follow.
AlleyCorp invested in Valar Atomics at a $20 million valuation less than three years ago. Ryan says its latest round, led by Sequoia, valued it at $6 billion.
While building DoubleClick he kept three things (work, family, fitness) and cut everything else by about 80%. He now spends 8 to 10 weeks a year out of the office and works a few hours on most of those days.
Michael Pollan’s How to Change Your Mind changed his view of psychedelics at 54. It takes $200 to $250 million to get one compound through the FDA, which is why he moved from donating to starting a company.
Transcend was set up as a public benefit corporation whose first shareholders pledged 10% of their gains. Ryan says that means about $20 million going to psychedelic causes over the next nine months, in roughly 20 large grants.
His billboard is about the country and not about startups: reduce the deficit, have the wealthy pay 30% in tax, fund retraining, and admit a million or two of the world’s most talented immigrants.

Chapters

2:40 Twenty Trends on a Whiteboard and the YouTube He Didn’t Start

Ryan lists the ten-year trends he has bet on: internet advertising, unstructured data, psychedelics, nuclear energy, the creator economy and value-based care. He admits there are trends, longevity among them, where he still has no product idea, which is one reason he runs four conferences and takes small groups on trips with assigned seating and a topic for each guest. The method is to ask who supplies the winner and how behavior changes. His example is falling bandwidth prices in 2003, which pointed straight at online video two years before YouTube launched.

13:45 Gilt Groupe: $500 Million in Year Four, Then a Falling Knife

Gilt brought the flash sale to the United States after Ryan saw Vente-Privée doing a billion dollars of it in France. Growth was extraordinary until designer brands and department stores built decent websites and began discounting their own stock. Gilt never became large enough to have a moat, so Ryan pushed the board to sell at a quarter of its peak valuation. He has stayed out of e-commerce since, on the view that delivery, price and returns leave a startup nothing to fix.

18:06 Dilbert, DoubleClick and “the Next Internet”

After investment banking, INSEAD and the launch of Euro Disney, Ryan became CFO of a United Media division and launched the Dilbert website in 1995. He sold its first banner ad to IBM at a price he made up on the spot. When he asked the parent company for a few million dollars to build an internet division, an executive told him they would wait for the next internet. Ryan left, joined DoubleClick as roughly its tenth employee and rose from CFO to CEO by deciding faster than competitors and expanding abroad before any country was profitable.

24:00 The AlleyCorp Model and Recruiting Henry Blodget

AlleyCorp is not an accelerator. Ryan and Dwight Merriman acted as co-founders, funded each idea themselves, launched a product and only then went to outside investors. Recruits had to believe two things: that the idea was great and that the pair added value. Henry Blodget agreed to run Business Insider after a 30-minute pitch built on a simple gap, which was that no business news site was native to the web. The site tested four headlines for five minutes each and never spent money on marketing.

31:34 The Two-Week Business Crush

Ryan and Merriman chose ideas by sitting in a conference room and talking, sometimes for months with nothing to show. He knows an idea is real when he cannot stop thinking about it for two weeks, and he compares it to knowing you have met the right person. He skips the spreadsheet because the one number that matters, how many people will use the thing, cannot be known in advance. Early on he wants a product person and no finance or marketing hire. Getting 100 million readers is 90 to 95% of the job, and selling the ads afterward is the easy part.

34:55 Valar Atomics and Why the Fund Stays Small

The nuclear bet came from AlleyCorp’s deep tech team at a time when fusion had the attention and the US government was not yet behind fission. Ryan credits the Trump administration for moving both nuclear and psychedelics forward, and points to France’s 50 years of running mostly on nuclear power. AlleyCorp has about 40 to 50 limited partners, with Ryan the largest. He keeps the fund small because the firm’s value is in a company’s first five years, and plans to sell roughly a third of a position four or five years in.

40:25 Ping Pong, Three Priorities and Ten Weeks Out of the Office

Ryan has played competitive table tennis since high school, still trains with a former Nigerian Olympian and plays with the Yale varsity team when he is in New Haven. During the DoubleClick years he protected work, family and fitness and dropped friends, culture and Sunday sports almost entirely. He says he has never burned out because he paced himself. A typical day away now is hard exercise in the morning in France, then a fixed three-hour block of New York meetings.

45:05 Methylone, Transcend and a Public Benefit Corp That Pays Out

Reading Michael Pollan led Ryan to Yale’s psychedelic researchers, first as a donor and then, in 2021, as co-founder of Transcend Therapeutics. The company chose methylone, a shorter-acting relative of MDMA, and patented it for PTSD, depression and anxiety. Ferriss’s opening notes say Otsuka completed its acquisition of Transcend in June 2026. Ryan points out that the largest source of PTSD is sexual assault, not combat, and that the trials were 60% women. The pledged 10% of early shareholders’ gains is now being given away, with interest in proving that group therapy can bring the cost down.

59:13 Deep Tech, Robots and MongoDB’s Revenue-Free Years

Asked whether early-stage venture is dead now that anyone can clone a consumer app, Ryan says the opportunity has moved to energy, robotics, space and healthcare, where the improvements are 10x and 100x. He describes an AI support agent that noticed a caller struggling, offered Spanish and resolved a medical question that English had stalled. He also covers a failed healthcare platform and MongoDB’s long road, comparing a new database to a new pacemaker that nobody wants to try first. He has no fixed rule for selling public shares and has sold MongoDB slowly over time.

1:08:44 The Billboard: Inequality, Taxes and Immigration

Ryan is not worried about American startups. He is worried about income inequality, falling mobility and the jobs AI will remove. His two levers are a fairer tax system that shrinks the deficit and serious retraining, plus many more skilled immigrants, a position he calls contrarian today. He compares the country to the Premier League, which is the best in the world because it takes the best players from anywhere. He ends on New York, which he says is booming, noting that MongoDB was run by an immigrant and Datadog was founded by French entrepreneurs.

Notable Quotes

“Startups have to solve a problem. Even if you don’t realize that problem is there, when you see it, you’re like, ‘Oh yeah, that is a problem I had.’”
Kevin Ryan, on why he has stayed out of e-commerce since Gilt

“No, because we’re going to wait for the next internet.”
A United Media executive, as Kevin Ryan recalls him turning down an internet division in 1996

“I start with a business crush, meaning there’s an idea and I can’t stop thinking about it. And then a week later, I’m still thinking about it.”
Kevin Ryan, on how he knows an idea is worth starting

“Big funds focus on the 2%. We focus on the 20%.”
Kevin Ryan, on why AlleyCorp does not raise a larger fund

“So, sometimes it’s better to be the second mouse.”
Kevin Ryan, on learning from the first psychedelic drug companies

“A database is like a pacemaker. If I said to you, I’ve got this startup pacemaker idea. Do you want to try it?”
Kevin Ryan, on why MongoDB took ten years to reach its IPO

“When families don’t feel like their kids are going to do better, they get desperate and they go to the far right or they go to the far left.”
Kevin Ryan, on what he would put on a billboard

Watch the full conversation between Kevin Ryan and Tim Ferriss here.