#884: How to Spot 10-Year Trends and Build Billion-Dollar Companies — Kevin Ryan of DoubleClick, MongoDB, Business Insider, and Many More

Why does it matter? Because billion-dollar bets are decided by a two-week gut test, not a spreadsheet.

Kevin Ryan has won in internet advertising, unstructured databases, luxury flash sales, financial media, and psychedelic therapy — completely different industries, same underlying operating method. The through-line isn't pattern-matching genius; it's a specific set of behavioral disciplines for identifying and riding 10-year trends before they get crowded. What this conversation unpacks is where that edge actually lives.

• The business crush test — two weeks of unshakeable obsession — predicts founder conviction better than any financial model
• Revenue without a moat is a trap: Gilt hit $500M in sales and still nearly became worthless
• DoubleClick dominated by operating in 25 countries before the first was profitable
• Consumer tech is largely solved; the remaining 100x improvement curves live in deep tech, robotics, and nuclear

The real opportunity is always one level downstream from the obvious trend

Most investors compete to own the trend. Ryan looks one step further.

"Let's imagine Shopify does very well. Who else works with them and supplies them? That's a second order." He applies the same lens to AI infrastructure: "The people providing the picks and shovels to data centers are doing very well."

His clearest second-order miss: in 2003, he tracked bandwidth prices falling toward a crossover with advertising CPMs. "The reason there was no video in 2003 is because it costs about $10 per thousand to serve content and you could get about a dollar in advertising." By 2005 or 2006, the lines would cross. He saw it, mapped it — and didn't start YouTube.

In robotics today, Ryan is asking which factory verticals will consolidate to single-use automation first, not who builds the best general-purpose machine. The general robot is the obvious bet. The textile factory deployment contract is the second order.

Stop one step further. That's where the uncrowded bets live.

Two weeks of obsession beats any financial model

Ryan's actual decision filter isn't a spreadsheet. It's a physical sensation.

"I'm 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." If the idea keeps compounding for two weeks — generating product visions, staffing scenarios, competitive angles — he acts. "If that sticks with me for like two weeks, I generally do it. Like I don't actually spend more time."

The Business Insider founding pitch took thirty minutes on one question: will people read business news online, and can a new entrant do it better? "If you don't know that, you don't know anything." No revenue projections. No market maps.

This isn't romanticism — it's a filter for genuine conviction. The two-week test weeds out ideas that are intellectually plausible but emotionally hollow. Founders who bail when execution gets hard, Ryan would argue, were never truly obsessed in the first place.

$500M in revenue, and Gilt still nearly became worthless

Four years in, Gilt Groupe was doing half a billion dollars annually in luxury flash sales. Then the market closed around it.

Brands like Marc Jacobs had no direct-to-consumer website in 2008 — wholesale businesses had no reason to build one. Gilt captured that gap entirely. By 2011 and 2012, those brands had their own sites. Department stores got their digital operations to mediocre. Farfetch arrived. "All of a sudden, I'm competing with a lot of people for that merchandise. It got a little commoditized, and we couldn't figure out a way around that."

Ryan pushed to sell when the company could still command $400 million. The board hesitated — they'd been worth a billion. He sold anyway, to Saks, for $250 million. Three years later, Saks offered to sell it back for $5 million.

The diagnosis: "Do you have a moat? We were trying to get to be big enough that we would have that moat. And we got to be quite big, but not big enough." Even at $500M revenue, Gilt could only buy 1,000 units from a vendor with 20,000 items to move — not enough to lock up supply.

Before you scale, know what keeps competitors out at 10x your current size. If the answer is unclear, the window you're in is a timer, not a business.

DoubleClick opened 25 countries before the first one was profitable — and that was the entire strategy

Twenty-five countries in three years. Competitors were in six.

Ryan made that call knowing none of the new markets had turned profitable. "We were in 20 countries before our first country was profitable. If it never worked, you'd be like, what were you doing?" The logic was enterprise-first: Microsoft and Procter & Gamble needed a global digital ad partner. Secure them, and smaller clients follow by default.

"When I went to Microsoft or Procter & Gamble, we do business in a lot of countries. We need to work with you. And then once you had all them, smaller players were like, well, Procter & Gamble and Microsoft work with you, so we'll work with you." Near-monopoly, locked in before competitors could react.

In winner-take-most markets, the cost of waiting for profitability before expanding almost always exceeds the cost of the expansion itself. The unprofitable footprint was the moat.

Gilt and Business Insider both won by starting with one embarrassingly small beachhead

Gilt launched with one flash sale per week — women's clothing only. One became two, two became five. Men's arrived roughly a year later, then kids, then travel, then home.

Business Insider launched with three journalists on a single beat: New York technology. "It's better to do one thing really well than a bad job on everything." A Wall Street vertical followed, then defense, retail, eventually 600 journalists across dozens of beats.

The distribution strategy was equally stark: no marketing, ever. "We're going to write stuff that's so good that eventually we'll have 100 million uniques. And everyone's like, that's not a strategy." That was exactly the strategy, and it worked.

Different business models, identical underlying discipline. Dominate the smallest defensible territory first. Expansion is something you earn — not something you plan for on day one.

Methadone beats MDMA precisely because it is weaker

The winning psychedelic therapy compound isn't the most powerful one. It's the one that fits how healthcare actually schedules and delivers treatment.

MDMA depletes serotonin enough that weekly dosing isn't viable. Methadone doesn't. "You can take methadone once a week in our trials. You actually can't take MDMA once a week. It's a little bit more toxic." Session duration also aligns with real clinical workflows — critical for FDA scalability arguments. "The duration of action is much more compatible with healthcare as we know it."

Ryan explicitly credits watching the early MDMA trials stumble. The FDA advisory committee couldn't standardize bundled psychotherapy. Six-hour sessions couldn't scale. "Early worm gets the bird, but the second mouse gets the cheese." Transcend built on that institutional knowledge, securing a patent covering PTSD, depression, and anxiety — three indications, 20 years of exclusivity.

When evaluating any emerging therapeutic category, filter not just for efficacy but for regulatory and logistical fit. The second entrant who learns from the pioneer's FDA failures often wins.

Consumer tech is solved; the 100x improvement curves are all in deep tech now

There are 164 soft drinks. "The odds of you coming up with one that stands out is pretty slim."

Same-day delivery, frictionless returns, low prices — the consumer problem is largely gone. "Startups have to solve a problem. I can get anything delivered to my house in like 27 seconds. I can return it. It's inexpensive. So I don't know how to do better." The problems worth solving are elsewhere. "We're seeing 10x and 100x improvements in robots, which will change the world. I don't see the same thing in e-commerce."

Nuclear, advanced materials, biotech: unsolved problems, exponential curves, thin competitive sets. Alicorp's deep tech allocation has grown faster than any other vertical over the past four or five years — not as strategic diversification, but as a direct read on where asymmetric returns still exist.

Three priorities and an 80% cut to everything else is why he has never burned out across 30 years

At DoubleClick's peak — pre-IPO, 1,500 employees, 25 countries — Ryan ran his personal life on exactly three things: family, work, and fitness. Sunday sports? Gone. Cultural events? Mostly gone.

"I could only do three things. I needed to be working hard, spending all the time I needed to with my family, and staying in shape. But everything else, I cut by like 80%."

He has never burned out. Eight to ten weeks out of the office per year. A preference for people who work 47 intense weeks over 52 mediocre ones. The operating principle isn't about finding balance — it's about making the deletion permanent. Subtraction is the prerequisite for output that compounds across decades.

The next era of asymmetric returns requires things that can't be vibed into existence

The common thread across Ryan's best current bets — nuclear reactors, psychedelic therapeutics, factory robotics — is that none of them can be replicated over a weekend by someone with a laptop and a free API key. Each requires regulatory navigation, physical infrastructure, and domain expertise that takes years to accumulate. Software eating the world is largely over. The builders who win the next 20 years will need a different toolkit entirely. Ryan has spent 30 years building exactly that. Most of the field is only now catching up.

---

Topics: venture capital, entrepreneurship, trend spotting, deep tech, psychedelics, nuclear energy, startup strategy, business building, MongoDB, DoubleClick, Business Insider, Gilt Group, Alicorp, methadone, second-order thinking, decision making