Rebecca Kaden

The through-line

Rebecca Kaden's consistent position is that the interesting object in venture is not the technology but the hypothesis it makes possible. "What interests me about technology isn't technology itself. It's the hypotheses about human behavior that it enables" [1]. Everything else follows from that: a thesis-driven firm that writes its arguments down and updates them, a preference for low-velocity, high-conviction investing, and a belief that returns come from ideas others do not yet hold rather than from chasing what is already moving. Her stated ambition is to be one of the best venture investors of her generation, and the thing she has conviction on is that "it is an impossible goal to meet without making Investments around ideas that are not obvious and others don't believe" [2]. Thesis-driven investing has been her declared subject across years of conversations, from USV's evolving thesis waves to the firm's more recent AI-era framings [4][10][11].

What has moved is the substance of the theses, not the method. She has gone from a consumer-focused early-stage investor working on brand building, defensibility and taking products from niche to mass market [3][5][8], to arguing about model layers, open-source stacks and programmable energy [1]. She is explicit that at USV "theses build on top of each other, they're never kind of on their own" [1], which is why an admission of uncertainty published roughly eighteen months ago became the foundation for a much more directional argument today.

On market cycles and how theses get built

She marks time in fund cycles and market cycles, and thinks constantly about how many of them she will get to see [1]. The framework she reaches for is Carlota Perez's installation and deployment eras: the internet had a massive installation era, then a boom and a bust, and out of the mess came the deployment era that produced Twitter and Etsy and the products that changed how people live [1]. Her read of the present is that installation has happened very fast over the last couple of years and the deployment cycle is bubbling up. "This is what a market cycle feels like. Let's go" [1]. She expects it to be messy and expects busts, and treats that as the price of the impact that follows.

The method is visible in a pair of published arguments. The earlier one, "the four quadrants," which USV also called the four futures, mapped outcomes against two axes: how open or closed the models are, and how fast progress continues before hitting asymptotes nobody currently foresees [1]. She notes how hard it is to write a piece whose punchline is that you do not know what will happen, and that LPs and founders both pushed for probabilities on each quadrant. USV refused. "Our punchline was we want exposure to all four quadrants" [1]. What the piece did do was identify what is advantaged in each world, which is how energy emerged as the horizontal bet. She also concedes what the framework got wrong in one direction: the fear that if the large models won, and especially if one of them won, "there would be no application era" and the application layer would be totally absorbed [1]. She frames AI as still dramatically underhyped, with abstractions built on top of the technology as what comes next [1][6].

On the rebel alliance and the application layer

The rebel alliance thesis is that a consortium can take on the big guys, and that a stack of pieces sitting on top of each other can take real market share [1]. Her historical argument is that the internet was built bottom up rather than top down, and that a top-down internet would have been neater but poorer: "the beauty of the internet are the weird corners" and the products that bubble up only because things are stacked together [1]. Today's market, she argues, is more top-down than it has been in a long time, a function of how capital has been deployed and of how large language models fundamentally work. She is careful not to overclaim: those models are working, USV benefits from them, and arguing otherwise would be strange. The claim is that there is also room for open-source models, open-source harnesses and tools usable without walled gardens around them, and that users and builders are both being pulled that way by control, flexibility and pricing power [1]. Asked whether that independence is exciting for a builder, she upgrades the framing: "I would say it's essential as a builder" [1].

Her renewed conviction in the application layer rests on two observations. First, there is more competition at the model layer than she once thought likely, because there is no single winner and because the model companies need to make money and so cannot keep subscription prices deeply subsidized. That is bad for consumers and good for startups, because "competition creates opportunity at the application layer" [1]. Second, and in her words something they should have known all along, "it's really hard for big companies to do all the things" [1]. The lesson repeats in every cycle: incumbents launch consumer verticals, some work and some do not, as was true at Google and Meta, and the pressure on the model companies to keep raising capital at scale may force even more focus [1]. Her broader ambition for the application layer is software built to obliterate markets rather than merely automate them [1].

On energy as the layer underneath everything

Energy is USV's biggest bet inside the four-futures framing, chosen precisely because it is advantaged in every quadrant: even if the large models win the day and own everything, the ecosystem still needs dramatically more, faster, and ideally cheaper energy [1]. She dates the firm's commitment to an energy fund launched around 2019, well before this became a consensus view, and describes exposure across the whole stack, from programmable energy, meaning the ability to control and direct energy in new ways with technology, to new formats of generation such as Radiant and Zanskar in geothermal [1].

On conviction, non-consensus and resisting FOMO

She distinguishes two investor skills and is clear about which she rates higher. There are investors who are excellent at picking up on momentum and running after it, a genuine skill she admires and aspires to, and there are investors who are experts at seeing around the corner, which she admires more [1]. The strategic argument is blunt: chasing the same deals as everyone else "is going to optimize prices you're never going to optimize returns that way" [2]. She rejects market share as the goal for a venture firm even while advising portfolio companies to build moats and pursue it, on the grounds that the muscle that produces excellent returns is decision-making and the ability to hold a hypothesis and invest behind it when others do not believe [2]. Her own preference, stated as a set of things she thinks are true about herself, is early stage, Series A, "low velocity High conviction" [2][9].

She does not pretend to be immune to the pull of hot markets. "I've never met anyone who doesn't experience a little bit of fomo" [2], and she names it as harder to resist earlier in a career. Her defense is the partnership itself: partners who say plainly that chasing is not how this group creates returns have been an invaluable counterbalance to a natural instinct [2]. The emotional register she is trying to learn from them is "high ambition low anxiety" [2]. She describes herself as sometimes high ambition and somewhat high anxiety, notes that the anxious version does not get you anywhere, and points out that an industry full of players who feed off your anxiety makes you a worse partner to founders [2].

On the small partnership as a deliberate constraint

USV has stayed consistent while peers grew AUM, and she accepts that characterization as accurate and intentional. The firm talks about "what's sacred in core and what's flexible on the edges," is not especially interested in structure, rules or process, and can only afford that looseness because the core is settled [2]. That core is a small, consensus-driven, equal, idea-driven partnership of five to six partners who co-run the firm, and that model simply does not scale past a certain size [2]. The reasoning is a mix of what maximizes returns and what the partners enjoy, and she likes that the honest answer to why the firm works this way is often just that they like it that way [2].

She is candid about the costs. Nobody knows who manages what or who manages whom, so someone has to raise a hand each time the annual meeting or a new fund comes around; the whole partnership once decided where to move the office, which she grants is not the most efficient process [2]. There is less leverage and fewer support functions than at other firms, so everything rests on the partners. Hiring is the hardest part, because the bar is extremely high and there are no junior investors working up the ranks, so there is no way to test someone [2]. What she thinks makes partnerships function rather than fall into the dysfunction venture is famous for is mutual respect and trust grounded in proven, differentiated strength: Fred as the firm's most public face with a very big voice largely because of his blog, Albert on certain questions, John on FCC matters, Andy on marketplaces [2]. That structure gives her the drive to prove her own domain and become that valued voice too. Underneath the partnership question is her view of what a firm even is: investing is "just a uh group group of people and a pool of capital," and since capital is largely commoditized, "it's just who you want to do this with" [2]. She rates the quality of the people as priority one through five, above stage or strategy preferences.

On the variety of venture models

She refuses to treat the platform-versus-craftsman debate as a contest with a right answer. Aggregation will always exist because some people find joy and success in making things bigger, and because the industry rewards AUM on near-term cash, which is simply a mechanic everyone deals with [2]. She does not think that is bad, just one way to optimize a market, and argues the industry would be far less interesting if everyone agreed on a single best model. Everyone benefits from the variety [2]. Choosing between models is, in her account, a matter of personality and inclination rather than correctness: she has never been a big-company person, is not good at politics and would not enjoy them, and prefers a small culture where everyone is in sync [2]. The taunt she enjoys most is a rival's line that a big platform firm is Beyoncé and USV is an indie band, meant as an insult and received as a compliment: "we love Indie bands" [2].

On whether venture can be automated

She is open to technology improving almost everything and to data providing inputs the industry does not currently have, in formats nobody can yet predict [2]. What she cannot see is automation replacing, particularly at early stage, the element of decision-making and conviction that is essential to investing: "I have never been able to figure out how that could happen um if anyone could I think it would be kind of incredible but it's not clear to me that's possible" [2]. She also notes that no firm has dominant market share at any stage, that the ecosystem is vast enough that a firm like USV could have incredible returns without ever looking at a YC company, and that dominance is not really the goal anyway [2].

On becoming an investor from outside

She came from a background with essentially no exposure to venture, felt like a real outsider, and entered with "imposter syndrome times a thousand," because the industry quickly demands that you have legs to stand on and explain what you offer [1]. What took time was learning on the job what her actual strengths were, rather than defaulting to the last thing she had done, and she is now grateful for the self-reflection that forced, because it is where the confidence came from [1]. She credits her journalism background with the instinct she uses to spot breakout companies, understanding how a story comes together before everyone else can see it [1][9]. The pull of the work is a treadmill of continuous learning against a job whose core is thinking about what is around the corner [1]. She describes the loop she wants to repeat: getting excited about an idea and a hypothesis, pulling that thread through false roads and right ones, and finally reaching the conversation with a founder where you say "You're who I've been looking for" [1]. She frames each commitment as a decade-long road and does not imagine getting off it, while also expecting her future to hold many other explorations and projects [1].

On consumer investing and getting an investor's attention

Her earlier body of work was consumer: identifying emerging consumer entrepreneurs, running seed programs, and the patterns of entrepreneurship behind taking a consumer product from niche to mass market [3]. Recurring subjects include leveraged growth, consumer brand building and customer obsession, worked through examples such as Allbirds and Zulily [5], plus defensibility in CPG and where the next generation of consumer companies comes from [8]. She has also addressed the earliest theses at USV, including its investment waves and its work in healthcare and fintech [10]. On the practical question of standing out in a flood of inbound pitches, she has argued for leading with a clear description of the venture and demonstrating alignment with the investor [7].

Takeaways

• Judge a technology by the hypotheses about human behavior it unlocks, not by the technology itself [1].
• USV declined to assign probabilities to its four futures and instead bought exposure to all four quadrants, with energy as the bet that pays off in every one [1].
• More competition at the model layer, driven by no single winner, rising prices and credible open-source options, is what restored conviction that the application layer will not be absorbed [1].
• Independence from a single foundation model provider is not a nice-to-have for builders, it is essential, and open stacks bring control, flexibility and pricing power [1].
• Out-chasing a market optimizes prices, never returns; excellent returns come from holding a hypothesis others do not believe [2].
• USV's small, equal, consensus-driven partnership is a deliberate cap on size, chosen for both performance and enjoyment, with real costs in hiring, leverage and firm-management decisions [2].
• Data will change venture inputs, but early-stage conviction and decision-making have no visible automation path [2].
• Since capital is largely commoditized, the choice of who you invest alongside outranks stage, strategy and firm model [2].
• Aim for high ambition and low anxiety; anxiety makes you a worse partner in an industry full of people who feed on it [2].