Kevin Ryan Looks for Businesses Created by the Second-Order Effects of Trends

Kevin Ryan starts with a long horizon. An important company, he argues, usually takes about a decade to build. That makes a trend that is already hyped—and may be over in two years—a poor starting point. The question is not simply whether a market will grow, but what that growth will make possible that does not exist yet.

Ryan describes writing down roughly 20 trends he believes could support large businesses over the next ten years, then asking what follows from each one. Internet advertising was the trend behind DoubleClick. Unstructured data in databases was the bet behind MongoDB. He cites psychedelics, nuclear energy, and the creator economy as other trends AlleyCorp backed, while noting that recognizing a trend does not guarantee he will find the right product. Longevity interests him, for example, but he says he has not yet found an idea he wants to pursue.

That distinction—between a trend and a product opportunity—is central to Ryan’s account. A large or growing market does not tell him what to build. He looks for an implication of the trend: a new behavior, a missing service, or a supplier that will matter if the obvious business succeeds. His example is Shopify. If Shopify grows, he asks, who else benefits by supplying it? Businesses providing “picks and shovels” to data centers offer a similar kind of opportunity: their prospects may follow from a larger shift without requiring them to build the most visible product in that shift.

He also asks how behavior might change—what people will stop doing and what they will do instead. These are not mechanical screening questions. Ryan says there is no formula that reliably turns the exercise into a company; it still involves instinct and guesswork. A trend can be clear while the right business remains unclear.

An early miss makes the second-order point concrete. In 2003, Ryan and his colleagues saw that bandwidth prices were falling. Serving video cost roughly $10 per thousand, while advertising brought in about $1. They thought the economics would cross by 2005 or 2006, but did not start a video company. YouTube launched in 2005. Ryan calls it a missed opportunity: they noticed the shift in economics but did not act on what it could enable.

His signal for acting is less a spreadsheet than persistence. He calls it a “business crush”: an idea that continues occupying his attention and keeps expanding in his mind. He imagines what could be built and whom he might bring in. If the idea still has that hold after about two weeks, he generally acts.

The initial work is not a detailed forecast of revenue or an exit. Ryan says he cannot know how many people will eventually read a new business-news site. The more basic questions are whether people will read business news online, whether the company can serve them better than established publishers, and whether it has a distinct vision for doing so. At the beginning, he wants a strong product—not a finance or marketing function. For Business Insider, he adds, attracting a large audience was much harder than selling advertising against it.

Ryan does not say that business models and capital do not matter. Rather, he does not want forecasts that cannot yet be known to substitute for deciding whether the product is compelling. He does not plan around exits, either. His stated priority is to build a product people want to use, then address financing, hiring, and growth as the company develops.

That preference also shapes what he considers a meaningful opportunity. A broad trend can be real and still not yield a startup. Ryan says he has not returned to e-commerce since Gilt Groupe because the problem he once saw there has largely been solved: consumers can get goods delivered quickly and return them cheaply. By contrast, he sees substantial work remaining in areas such as cancer treatment, nuclear energy, and mental health.

Ryan also uses events to learn about trends and people in fields he does not yet know well. AlleyCorp runs four gatherings, including Deep Tech New York, Digital Health New York, DOC, which focuses on longevity and science, and Odyssey, which is less commercial and brings people together for intellectual exchange. He describes Odyssey as nourishment rather than a direct business-generation exercise. If he meets fascinating people working in different areas, he believes useful things can follow, whether or not they lead to a company.

The format is designed to make those conversations happen. A hiking gathering might bring around 30 people together for a three- or four-day trip: a four- or five-hour hike in the morning, sessions on topics such as climate or nuclear energy in the afternoon, and dinner afterward. Ryan assigns seats during the first two nights so people mix, then leaves the last night open for people to reconnect with those they most want to see. Smaller trips can be more structured: on an annual biking trip of eight people, each person leads one 20-minute conversation over a meal. Subjects have ranged from crypto to how well-off parents talk to their children about money.

These events are not all networking exercises in disguise. Ryan says some are linked to his businesses, while Odyssey is less commercial and serves as intellectual nourishment. He also organizes trips around the chance to spend time with people from different fields. For a planned trip to China, the question is what the country does better; he says he wants to see that for himself because he has not been there in a long time and does little business there.

Kevin Ryan saw Gilt Groupe as a way to take a sample-sale experience beyond the cities where it happened. He had spent time in France, where Vente-privee was already running online flash sales of discounted luxury goods. In New York, he remembered seeing a long line outside a sample sale and thinking about people in places such as Philadelphia or Austin who could not easily access the same merchandise. Gilt could bring those sales to them.

The retail interface made the offer tangible: customers could see discounted designer goods, select a size, add an item to a cart, or join a wait list when a product sold out. But the company was not simply matching buyers with merchandise. Ryan says Gilt had to buy, package, sell, and handle returns for the goods. The service solved an access problem while taking on the operational work of selling physical products.

The business grew quickly: Gilt recorded $175 million in revenue in its second year and $500 million in its fourth. But its early advantage depended on a gap in the market. Luxury brands and department stores did not yet sell effectively online or discount their own merchandise there. As brands built their own websites and department stores improved theirs, Gilt had to compete for the same inventory with more players. Ryan says the merchandise became commoditized, and the company could not find a way around it.

He had hoped scale would create bargaining power with suppliers. It did, but not enough. One brand, Theory, had 20,000 end-of-season items to sell; Gilt could buy 1,000. That was a useful amount, Ryan says, but it did not move the needle for the supplier. Gilt was large, but not large enough to shape the market in the way he had expected.

At one point, Gilt had been valued at about $1 billion. Ryan told the board he thought the company should be sold, even if the likely price—perhaps $400 million—was disappointing. He saw the business as a “falling knife.” The eventual sale to Saks was for $250 million; three years later, Ryan says, the buyer offered to sell it back for $5 million.

The sale was not a great outcome, in his view, but it was better than waiting as the business deteriorated. The experience sharpened the distinction between growth and defensibility. A company can be large and still lack the purchasing leverage or other protection needed to withstand competitors copying its model. Ryan says Gilt “almost got there but didn’t.” Its scale helped it buy inventory, but its orders remained too small to give it influence with some suppliers. The gap it had filled also narrowed when brands and department stores improved their own online sales.

The same discipline appears in AlleyCorp’s approach to starting companies: begin narrow, prove the product, and expand as the business earns the right to do so. Gilt started with one women’s clothing sale a week, then moved to two, three, and five. About a year later, it added men’s clothing, then children’s products, travel, and home. That progression let the company add categories over time rather than trying to serve every possible customer from the start.

Business Insider followed a similar path. It started with three people covering New York technology, a small beat the team believed it could cover well. It then added a Wall Street vertical, followed by more areas as the newsroom grew. At its peak, Ryan says, it had about 600 journalists, including six covering defense. The point was to do one thing well enough to attract readers, then broaden the operation as it grew.

Business Insider’s original premise was that financial news had not been built for the web. The Wall Street Journal and Businessweek did not update their websites throughout the day. Ryan says Business Insider published multiple stories as information developed, used punchier headlines, and tested different headlines to see which drew readers. At the time, testing four headlines for five minutes each was unusual. It made sense online, he argues, because a headline could affect readership in a way it did not for a magazine article.

The company avoided marketing and relied on the content to bring in an audience. Ryan recalls that people doubted the plan to write well enough to reach 100 million unique visitors. He says that is what happened. For him, acquiring that audience was far harder than selling advertising against it: with 100 million unique visitors, he says, a company could hire someone to sell ads. The audience was 90 to 95 percent of the work.

Kevin Ryan and technical partner Dwight Merriman supplied the initial capital for AlleyCorp’s early companies and acted as co-founders. This was not a program for outside founders to bring in ideas and pitch at a demo day. They usually spent about a year building and launching a product before raising venture money. In the first phase they started six companies, in two groups of three; Ryan says three became hugely successful: Gilt, Business Insider, and MongoDB.

Their roles were complementary. Merriman, whom Ryan describes as one of the strongest technical minds he has worked with, handled the technology. Ryan concentrated on the business side: finding a CEO, building the company, and raising money. Their earlier success at DoubleClick helped recruit people. They had just sold what Ryan describes as the most valuable startup created in New York at the time, giving candidates reason to believe the pair could help them build and finance a company.

The pitch to a potential CEO had two parts: believe the idea is good, and believe Ryan and Merriman can add value. Henry Blodget is one example. Ryan knew him somewhat and brought him the Business Insider idea. After about 30 minutes, Blodget decided it was worth pursuing. Ryan calls him an exceptionally talented writer. Other people who became significant journalists joined early in their careers, but recruiting them took years, not weeks, and depended on the company’s growth.

Ryan does not try to predict in advance which venture firms will fund a company. When AlleyCorp raises, he says, it may approach 20 or 25 firms. The investor he expects to be interested is often not; someone else falls in love with the idea. He does not want to pre-set that choice. The work before fundraising is to build and launch a product, then see who responds.

Business Insider’s early expansion and DoubleClick’s international push show two different ways Ryan has tried to grow from a narrow starting point. Business Insider added coverage as it gained readers and journalists. DoubleClick made a faster, riskier bet on multinational reach. Ryan says the company opened offices in 25 countries in its first three years, before any single country was profitable. That footprint helped win multinational clients such as Microsoft and Procter & Gamble. Once those companies were working with DoubleClick, smaller players had another reason to work with it too. Competitors were present in only six countries, Ryan says, and the global reach helped DoubleClick dominate its space.

The expansion meant accepting a period in which the local economics had not yet worked. Ryan says they were in 20 countries before the first country was profitable; if the strategy had failed, the decision would have looked questionable. He attributes the willingness to expand to making decisions faster than competitors, understanding the numbers, and taking risks. DoubleClick is an example of a bet that multinational customers would value a footprint competitors could not yet offer, not a rule that entering many markets early is always right.

The contrast with MongoDB was stark. The database company had no revenue for three and a half years. Ryan says it was a hard product to sell, comparing it to a pacemaker: a customer may not want to try an unproven system, but without early users it is difficult to establish a record. MongoDB gave the product away for a long time and found inexpensive use cases before it could build trust and revenue. Eight years in, it was doing about $40 million in revenue—modest by the standards of its later value, Ryan says, but with usage growing and the product improving. It went public ten years after founding.

Ryan’s contrasting examples resist a simple playbook. DoubleClick went public 24 months after it started; MongoDB took a decade. The first had a sales case for broad international reach; the second had to build adoption for a difficult product over time. Ryan says that if people are taking notes, they should not imitate MongoDB’s years without revenue. He also says the product worked and improved, and that usage was growing even when revenue remained modest.

AlleyCorp’s own failures are less easily explained. The team noticed that digital-health providers were building similar online features, including scheduling and patient information, and that the systems had to meet HIPAA requirements. It tried to create a kind of Shopify for healthcare sites. The company raised $6 million, but providers were reluctant to sign up, and AlleyCorp eventually closed it.

Ryan cannot say whether the company needed better execution, more time to win its first ten prominent clients, or something else. He also notes that healthcare companies were reluctant to outsource to a new provider. The repeated features seemed to point to a product opportunity, but that did not resolve customers’ hesitation to trust a new supplier. Ryan says the team still does not know exactly why the idea failed.

The path that led Ryan to DoubleClick began outside the internet industry. After studying economics and working in finance, he moved into operations at Euro Disney, where the launch involved 15,000 hotel rooms and 50,000 people a day. He says he learned from the scale of the project, but decided he did not want to work for large companies. Back in New York, he became CFO and COO of a 180-person United Media division, managing operations and finance. That role gave him a larger operating remit, but it was the internet that changed his direction.

Ryan remembers reading a 1994 Businessweek article about the internet and thinking it could let people communicate and buy things in a new way. In 1995, he launched the Dilbert website because United Media owned the comic’s intellectual property. The site attracted substantial traffic from people who could use the internet before browsers were widely used. Ryan began selling advertising, setting a price on the spot for IBM, which bought the first ad for two weeks. The team also sold merchandise such as T-shirts and ties.

After a year, Ryan asked United Media to give him a couple of million dollars to build an internet division serving other companies. He says the executive he approached declined, saying they would “wait for the next internet.” Ryan took that as a sign that he should leave, not as evidence against the business opportunity. He says the internet was already gaining users, and people were beginning to shop and use maps online. He decided to start an internet company and joined DoubleClick as roughly its tenth or twelfth employee, first as CFO, then president, and eventually CEO.

Kevin Ryan says AlleyCorp now organizes its work around three areas: enterprise and consumer technology, healthcare, and deep tech and robotics. The firm’s portfolio includes companies such as MongoDB and Business Insider in enterprise and consumer technology, healthcare companies including Thyme Care and Pearl Health, and Valar Atomics among its deep-tech investments. Ryan says deep tech has grown the most over the last four or five years. He sees large markets in energy and automation, alongside businesses created by the infrastructure those industries require.

Valar Atomics illustrates both the scale of the opportunity and the uncertainty of forecasting it. AlleyCorp invested less than three years before the interview, at a $20 million valuation, after its deep-tech team brought the company forward. Ryan says the investment was based on the team and its ability to execute. The company later announced a large financing led by Sequoia; coverage Ferriss referred to reported a $6 billion valuation. Ryan says the company had been performing exceptionally well.

At the time of AlleyCorp’s investment, Ryan says, nuclear was not the obvious “hot” category; fusion had drawn more attention. He argues that advanced nuclear should be part of the energy future and says government support has since moved the field forward. For him, the thesis is not only about an individual reactor company. Energy demand, including the needs of data centers, creates opportunities across the industry.

The same downstream logic informs his view of space. AlleyCorp invested in Portal, a company that helps move satellites from one area to another. Ryan sees a wider ecosystem forming around the expected increase in satellites: the opportunity may lie not only in building satellites, but in the services and infrastructure required when there are many more of them.

In robotics, Ryan expects the number of robots to increase dramatically, but sees the strongest early business cases in specific tasks and controlled environments. A factory can be designed around a robot that repeats one job well. A restaurant worker, by contrast, may need to chop vegetables, carry waste up stairs, and fetch something from a nearby shop. That range of unpredictable tasks is harder to automate. Ryan expects robotics to spread through factories, with highly specific applications rather than a single general-purpose machine doing everything.

The same focus on a discrete unmet need underlies AlleyCorp’s investment in Aescape, a robotic massage system. Ryan says there were 17,000 unfilled massage jobs in the United States two years earlier. A hotel guest who cannot book a human therapist at a convenient time might use a machine; some customers may also prefer a robot because they are uncomfortable undressing around a person. Ryan says the product is still evolving, but estimates it is about 70 percent of the way to solving the problem.

Not every automation opportunity is settled. Ryan says the restaurant problem remains difficult. Customer support, however, has already produced an example that impressed him: an AI agent on a medical call noticed after 15 seconds that the caller might be more comfortable in another language. When the caller agreed, the agent continued in Spanish and resolved the issue. Ryan says the caller had struggled to explain the problem in English; switching languages changed the call.

These examples show how Ryan applies the search for implications of a broader trend. “More robots” or “more AI” is not yet a company thesis. The practical question is where a capability can perform a specific task reliably, or where a new industry will create a supporting market that is not yet well served.

Kevin Ryan says his interest in psychedelics began as an intellectual change rather than an investment plan. At 54, he read Michael Pollan’s How to Change Your Mind, having never tried psychedelics or thought much about them. He says the book persuaded him that these medicines could be useful for people, and that academic work had already shown potential in areas including PTSD, depression, and anxiety. His work with Yale and its psychedelic research program deepened that conviction.

In 2021, Ryan and Yale professor Ben Kelmendi began considering a for-profit company as the next step. Ryan says taking a compound through the FDA process costs roughly $200 million to $250 million, more than a nonprofit could realistically raise for a single treatment. The team chose methylone, which Ryan says had preliminary results that made them think it could have an impact, while remaining unfamiliar to many people.

Ferriss describes methylone as a gentler, shorter-lasting relative of MDMA. Ryan agrees with the comparison and says the company saw potential advantages over MDMA. He says methylone can be taken once a week, while contrasting that with MDMA, which he says cannot be taken weekly, is somewhat more toxic, and depletes serotonin more. Ryan also says some people do not experience the “downer” after methylone that they can experience after MDMA. These are Ryan’s account of the compounds and the rationale for the company’s choice; he does not specify in this exchange that the weekly comparison refers to Transcend’s trials.

Ryan says the company began the development process and that the results “ended up to be incredible,” without giving trial details in this exchange. Transcend began with PTSD as its first indication, despite having patents covering depression and anxiety too. Ryan says each indication carries a substantial development cost, so the company could not pursue all three at once. The decision also reflected the difficulty of raising money during a period when psychedelic companies were struggling to attract capital.

The company’s timing and design were informed by what Ryan calls the “second mouse gets the cheese.” He and Ferriss discuss the risks of being first in a new treatment category: early companies can encounter regulatory or clinical problems that later entrants learn from. Ferriss points to challenges in the broader field, including the cost and staffing involved in administering long psychedelic sessions and the difficulty of standardizing psychotherapy when it is bundled with a drug treatment. Ryan says Transcend tried to learn from what it saw in the field and change certain things.

Transcend was structured as a public benefit corporation. Ryan says the initial shareholders committed to give 10 percent of their gains to a philanthropic effort, and that he and CEO Blake Mandell would decide where the money went. In July 2024, Transcend announced a $20 million initiative for research and access to psychedelic mental-health care. Ryan says in the later interview that the plan was to distribute $20 million over the following nine months to causes connected to psychedelics.

The allocation was still being worked out at the time of the interview. Ryan says the team was considering a relatively small number of substantial grants rather than hundreds of small ones, and looking for areas where funding could make a difference others were not addressing. He says he did not expect to give as much to pure academic research. One possible focus was access: Ryan says fewer than 0.5 percent of people with depression or PTSD had been able to access psychedelic treatment. He was interested in lowering costs and in testing whether group therapy could work as well as, or better than, individual treatment in some cases. He points to the support people get from others in group settings, while noting that the FDA has been nervous about how group treatment would be monitored.

The effort was not described as an open-ended grant competition. Ryan says a person was assembling information about the field and meeting with organizations; the team had identified around 50 organizations to consider. His stated preference was to make about 20 substantial grants, rather than 500 small ones. He expected the grants to support causes connected to psychedelics, but said they were still deciding where the money could have the greatest effect.

Ryan says the commitment served two purposes. He considered it fair, and wanted people joining the company to understand that its mission was part of the business. Employees, board members, and shareholders knew that the initial shareholders had committed to give away part of their gains. The arrangement, he says, was meant to remind the company that a possible financial exit did not replace its goal of helping people with PTSD and, eventually, other conditions. Ryan says he hopes others in the field follow suit, though he does not know how to require them to do so.

Kevin Ryan says AlleyCorp has 23 or 24 full-time employees and a relatively small group of limited partners, perhaps 40 or 50, with Ryan himself the largest. He says the first fund would be roughly at a 60 percent IRR at the end of that quarter. He also says his family office was at 60 percent, and the new $335 million fund was in that ballpark. Across the three funds he described, he said none had so far been below 50 percent IRR, while calling the results fortunate and crediting strong companies.

He does not want to scale simply because he could raise a larger fund. AlleyCorp, he says, is most useful in the first five years of a company, when it can help build rather than merely supply capital. He characterizes venture funds as earning “2% and 20%”: the management fee and the share of investment gains. Larger funds, in his telling, can become focused on the 2 percent; AlleyCorp wants to focus on the 20 percent.

That does not mean the firm will never grow. Ryan says its funds may increase in size as it adds investment areas, time, or perhaps a new geography. But he wants that growth to follow the work the firm does well, rather than a drive to accumulate assets. He describes building companies as both the firm’s strength and the reason he wants to keep doing it.

Ryan’s decisions about liquidity follow a similar practical logic rather than a universal rule. He says that for a smaller fund like AlleyCorp, it can make sense in rounds four or five years after investing to sell a third of the position. He says the firm expects to do that over the next year, returning some money to investors and teams while retaining exposure to future growth. He does not describe this as a rule for every investment. For public companies, he has no fixed policy. He held MongoDB through its first one or two years as a public company, then sold gradually as its share price rose, while retaining some shares.

His approach to time outside work has also involved explicit tradeoffs. During the intense DoubleClick years, Ryan says he recognized that he could focus on only three things: work, family, and staying in shape. He took his children to school and says he cut everything else by about 80 percent. That meant giving up much of the cultural life and socializing he had previously enjoyed, along with watching sports on Sunday afternoons. He says he paced himself and had not burned out.

Ryan wanted to avoid reaching the point when his children left for college and feeling he had not spent enough time with them. He says the family took at least four or five weeks of vacation together and made good memories. After his children went to college, he had room to add other interests. Now he spends eight to ten weeks out of the office each year, but says he still works daily, sometimes for only two or three hours and sometimes for a full day.

That time away is planned around work rather than treated as a complete disconnection. While staying at his house in France, Ryan might reserve 9 a.m. to noon New York time for meetings, then bike, play tennis, or swim. One day might include a longer work shift; when skiing, he says he might return by 3 p.m. and work until the evening. His case for taking vacation is not that work disappears, but that time with family and time to stay in shape need to be protected alongside an intense job.

Kevin Ryan says he worries about income inequality and declining chances for people to move up the economic ladder. He wants the United States to be more balanced, open, and caring. People face poor schools and insufficient support, he says, while the country is not well prepared for jobs lost to AI. He argues that these pressures could catch up with the country over time.

His proposed responses include a more equitable tax system and a lower budget deficit. Ryan says wealthy people should pay 30 percent in taxes, arguing that reducing the deficit could also reduce interest rates. He also calls for better access to education and professional training—not necessarily a college degree, but training that helps people find work and adapt. He says families need to feel their children can do better; when they do not, people become desperate and gravitate toward the far right or far left, neither of which he sees as the answer.

Ryan also argues for more immigration, particularly of highly skilled workers. He distinguishes between unlimited border crossings and recruiting people with advanced skills in fields such as AI and computer science. He sees little sense in blocking people who could join a company, earn a living, and later start businesses that employ others. The United States, he says, should welcome the most talented people in the world. He points to immigrants from Haiti, Nigeria, and Kosovo as people who have come to work hard, and says the country should attract more of them.

He objects to the message that immigrants are bad, and argues that companies in Silicon Valley have benefited from backing immigrants. His point is not that the country should accept unlimited numbers of people, but that it should make it possible for highly qualified workers to come. Ryan says he has seen pressure on academics’ ability to come to the United States, and says some academics have left for Oxford, Cambridge, or Canada. He compares the issue to English soccer: in his view, the Premier League is the best league because it draws players from many countries, not only England.

That view connects to his confidence in New York and the country’s technology sector. Ryan says New York is drawing entrepreneurs from Europe and points to MongoDB and Datadog—both founded by immigrants—as examples of major technology companies built there. He says New York is thriving and expects it to remain successful. More broadly, he argues that the United States is performing strongly in business and technology, while also saying the country should pay attention to what China is doing.

Ryan’s optimism about company-building does not, in his account, settle questions about who benefits from economic change. He sees room for companies to build products, enter new markets, and make treatments more accessible. He also says those developments do not answer whether people have good schools, training, or a reasonable chance to improve their lives. Company creation is one part of the picture he describes, not a substitute for the broader public choices he wants the country to make.