When we named this fund Swarm Capital, we were not being metaphorical. The swarm is our actual investment thesis.
In biological systems, swarm intelligence refers to the collective behavior of decentralized, self-organized systems — ant colonies, bird flocks, schools of fish — where the aggregate behavior of many simple agents produces remarkably sophisticated outcomes that no individual agent could achieve. The colony knows where the food is. The flock avoids the predator. The school navigates the current. Not because any individual organism has a complete map of the environment, but because information flows through the network, aggregates in ways that no single node can replicate, and produces decisions that consistently outperform individual intelligence.
We believe that a well-constructed venture portfolio works the same way. The best venture funds are not just collections of bets — they are interconnected networks where information, relationships, customers, and talent flow between portfolio companies in ways that create compounding advantages for the whole system. This belief shapes everything about how we construct and manage our portfolio.
The Thesis Behind the Portfolio
Swarm Capital's thesis begins with a simple observation: the most important technological and economic opportunities of the next decade sit at the intersection of three megatrends that are reinforcing each other simultaneously.
The first megatrend is the maturation of artificial intelligence from a research phenomenon to a platform technology. We are past the point of debating whether AI is real or overhyped. The question now is: which specific AI applications will generate durable value for which customers at what price point, and which teams are best positioned to build them? This is an empirical question, and it drives our conviction-based sector focus on AI infrastructure, vertical AI applications, and AI-enabled automation of knowledge work.
The second megatrend is the re-platforming of financial services. The financial system was built for a 20th-century world: slow, paper-based, relationship-dependent, and deeply underserving of anyone who did not fit the median customer profile of a large bank. The digital economy — global, instantaneous, diverse, and data-rich — demands financial infrastructure designed for it. Fintech is not a niche; it is the modernization of one of the largest sectors of the global economy.
The third megatrend is the emergence of software as the primary driver of enterprise productivity. The penetration of SaaS into the enterprise market is still increasing, and the next generation of enterprise software — contextual, intelligent, agentic, and deeply integrated into workflows — will generate returns that make the first wave of cloud SaaS look modest. Enterprise SaaS at the seed stage remains one of the most attractive risk-reward profiles in venture investing, precisely because so few investors are willing to back companies before the revenue profile becomes legible.
Why Portfolio Architecture Matters More Than Individual Picks
A common question I receive from potential LPs is: "What is your edge in picking individual companies?" My answer usually surprises them: individual company selection, while important, is not the primary source of returns for a well-run seed fund. Portfolio architecture is.
Here is why. The power law of venture returns means that the distribution of outcomes in a seed portfolio is extremely fat-tailed: a small number of investments generate the vast majority of returns, and even the best seed investors cannot reliably identify in advance which companies will be the outliers. The implication is that portfolio construction — how many companies, in which sectors, at what check size, with what follow-on strategy — is at least as important as the quality of individual investment decisions.
Swarm Capital has made deliberate choices about each of these dimensions. We invest in 18-22 companies per fund, which is fewer than many seed funds. We are concentrated in three sectors rather than investing across all of technology. We write initial checks in the $500K to $2M range and reserve 60% of our fund capital for follow-on investments in the companies that are executing. These choices reflect specific views about where we add the most value and where our portfolio architecture can generate outcomes that exceed what random selection across a broader portfolio would produce.
The Network Effects Within a Portfolio
The most distinctive element of Swarm Capital's portfolio strategy is what we call intra-portfolio network effects: the deliberate cultivation of relationships between portfolio companies that create compounding value for the whole swarm.
Concretely, this means several things. When we invest in a B2B SaaS company, one of the first calls we make is to our portfolio companies to ask whether the new company's product solves a problem they face. If the answer is yes, we facilitate a warm introduction and support a pilot or contract. This creates initial customers for the new company at zero cost of acquisition and creates real commercial relationships between portfolio companies that benefit both parties.
We have seen this play out in our portfolio with meaningful financial results. Three of our portfolio companies have signed enterprise contracts with other Swarm portfolio companies that now represent their largest revenue sources. Two portfolio companies have hired key executives who were referred through the Swarm network. One portfolio company completed a technical integration with another that reduced their combined infrastructure costs by 30%.
These outcomes are not accidental. They require active facilitation — regular portfolio summits where founders share challenges and opportunities, systematic introductions by the platform team, and a culture of reciprocity that we explicitly cultivate from the moment of first investment. But when the network effects compound, they create a genuine competitive advantage for our portfolio companies relative to those backed by funds that treat portfolio companies as isolated bets.
How We Evaluate Portfolio Balance
Beyond the intra-portfolio network effects, we think carefully about portfolio balance across several dimensions.
Sector balance within our thesis. Our three investment pillars — AI, Fintech, and Enterprise SaaS — are not independent. AI-native fintech is a category we have significant conviction in. AI-powered SaaS is where much of the enterprise software opportunity lies. We construct our portfolio to reflect both pure-play concentration and cross-sector opportunities that sit at the intersections.
Stage balance within seed. The seed stage itself covers a wide range: from pre-product companies with only a founding team and a thesis to post-seed companies with $2-3M ARR preparing for a Series A. We invest across this range, but we weight toward earlier-stage companies because the entry price differential and the impact of our operational support are both greater at earlier stages. Roughly 60% of our investments are in pre-product or pre-revenue companies; 40% are in companies with initial commercial traction.
Founder diversity. We actively seek to build a portfolio that is diverse along multiple dimensions: geography (not just SF and NYC), background (operators, domain experts, and technical founders), and demographic diversity. This is not primarily about social responsibility, though we believe it is the right thing to do. It is about return maximization: diverse teams consistently outperform homogeneous teams on the dimensions that matter most in company building, and we systematically look in places that less intentional investors overlook.
The Concentration Debate
One of the most actively debated questions in venture capital is concentration versus diversification. Some of the most successful venture funds have been highly concentrated — 10 to 15 investments per fund — while others have achieved strong returns through broad diversification across 100 or more companies. Where does Swarm Capital sit on this spectrum, and why?
Our answer is informed by the specific operational model we have built. We can provide meaningful support to 18-22 companies in a fund. The marginal value of adding a 23rd company to the portfolio is negative if it means diluting the attention we can give to the 22 companies already in the portfolio. This calculus is specific to our model and would not apply to a fund that takes a more passive, diversified approach to seed investing.
We also believe that our ability to add value — through portfolio network effects, operational support, and follow-on conviction — means that our portfolio companies have higher probabilities of reaching the next stage of growth than the base rate for seed-stage companies. If we are right about this, the correct portfolio size is smaller than the optimal size for a fund that cannot improve the probability distribution of outcomes.
Over time, the data will tell us whether this belief is justified. Three years into Swarm Capital Fund I, our preliminary data supports it: our portfolio companies are raising follow-on rounds at a rate that exceeds the seed-stage industry benchmark, and the average valuation step-up at Series A is meaningfully above the market average. This is preliminary evidence, not conclusive proof, but it reinforces our conviction in the model.
The Long View: What We Are Building Toward
Swarm Capital is not just trying to generate strong returns for our first fund. We are building a firm that will be the defining seed-stage investor in AI, Fintech, and Enterprise SaaS for the next decade. That ambition shapes decisions that might otherwise seem puzzling in the short term.
We decline investments in companies that could generate strong returns but do not fit our thesis, because building sector depth and reputation requires consistency over time. We invest in building our platform team before the returns from Fund I are clear, because the operational model requires the platform team to be fully functional from the first investment in each fund. We actively cultivate relationships with founders who are not ready to raise yet, because the best seed investments often come from multi-year relationships rather than competitive processes.
The swarm metaphor remains instructive here. A swarm achieves its remarkable emergent intelligence not through any single interaction but through the consistent application of simple rules over many interactions over long periods of time. Our firm is building its collective intelligence the same way: through consistent application of our investment principles, sector focus, and operational model across many interactions with many founders over many years.
The portfolio we are building reflects this philosophy. Each investment is both a bet on a specific company and a node in a network that we believe will generate compounding value for every company in the portfolio. Collective intelligence, exponential returns.
