Why We Invest in Volume
A few years ago I was sitting across from a seasoned LP, the kind who has backed dozens of managers, seen multiple cycles, and has very little patience for slides. He leaned back and said, “I’ve never seen a scatter-shot investor beat a focused one.” Then he paused. “But I’ve never seen a focused seed investor who could actually prove he was focused on the right things.”
That tension is real, and it sits at the heart of how Team Ignite Ventures thinks about portfolio construction.
We invest broadly at the pre-seed and seed stages. More companies than most firms our size. More first checks, more initial relationships, more early exposure. And LPs who hear that often have the same quiet worry: does this produce venture-scale returns, or does it just produce a well-diversified mediocre outcome?
The answer matters. So let’s work through it carefully.
The structure of how seed funds actually win
In 2020, two data scientists at AngelList — Nigel Koh and Abraham Othman — published a study of more than 10,000 LP investors on their platform. Not models. Not simulations. Actual investor portfolios, tracked over time, sorted by how many companies each investor held exposure to.
The results were striking. Investors with more than 50 companies in their portfolio earned a median annual IRR of 11.9%. Those with 50 or fewer: 2.9%. That gap held at every threshold they tested — 5, 10, 25, 50, 100 companies — and the direction never reversed. More exposure, better returns, consistently.
The “in the money” data was even cleaner. Nearly 90% of investors holding 90 or more positions had a portfolio worth more than they put in. Among those with three or fewer investments, fewer than half did. Same platform, same era, same deal flow available to everyone.
This is not an argument for mindless diversification. It’s an argument about the structure of how early-stage returns actually work. Venture outcomes are not evenly distributed. A small number of companies generate the overwhelming majority of value, and they are almost impossible to identify in advance. Ben Evans, a partner at Andreessen Horowitz, documented this precisely: across Horsley Bridge’s more than 7,000 investments made between 1985 and 2014, just 6% of deals produced a return of 10x or more. Those deals accounted for 60% of total returns. Everything else was noise around the edges.
AngelList took this a step further. Analyzing 1,808 early-stage investments across their platform, they ran simulations of what happens when you build concentrated portfolios — ten companies at a time — drawn from that universe. The most common outcome from a ten-investment portfolio was modest positive performance, well below the market return. The reason is mechanical: the typical concentrated manager simply doesn’t draw one of the outliers in ten chances, while an index across the full universe is guaranteed to hold all of them. Their conclusion was direct — a broad indexing strategy across credible early-stage deals would outperform roughly three-quarters of early-stage venture funds.
That’s the baseline. A passive index, no selection, no judgment, beats 74% of funds that are trying. Which means the question for any disciplined active manager isn’t whether volume helps — the data says it does — but whether their selection process and follow-on discipline can extract more than what the index delivers on its own.
We ran our own simulation to understand what portfolio size actually does to fund outcomes, drawing on the same power-law return distribution the academic research describes. The findings were striking in a specific way: the mean return barely changed as we added companies, because in a power-law world the rarest winners dominate the average no matter how many positions you hold. But the median outcome — what a typical fund actually delivers, vintage after vintage — climbed sharply and steadily. A 300-company portfolio produced median outcomes several times higher than a 50-company one. And the probability of holding at least one 1,000x company jumped from roughly 10% at 50 positions to better than 50/50 at 300. That is not a marginal difference. It is the difference between hoping and designing.
If the winners are that concentrated, and that hard to predict, then the most important question in portfolio construction isn’t “how much conviction do we have?” It’s “did we have exposure to the companies that mattered?”
A fund that concentrates early in ten companies and misses every one of the breakouts in its cohort doesn’t win on conviction. It loses on access. A fund that holds positions in 300 companies has dramatically increased its probability of being in the room when something extraordinary happens.
What “high volume” actually means in practice
There’s a version of broad-portfolio investing that deserves skepticism. If you’re writing checks without discipline and treating seed investing like a passive index, you’ve capped your upside without any mechanism to concentrate into the winners. That’s not a strategy — it’s abdication dressed up as diversification.
The version that works is different.
You write many initial checks, at attractive entry prices, against a structured sourcing and evaluation process. You do enough work upfront to separate the companies most likely to survive into genuine contention. Then, as the portfolio matures and real signals emerge — revenue growth, team execution, market traction, follow-on financing from credible investors — you concentrate your remaining capital into the subset of companies showing breakout trajectories.
This is not spray-and-pray. It’s a two-stage system. Broad exposure in stage one. Concentrated conviction in stage two. The first stage improves your odds of holding a winning position. The second stage amplifies the size of that position before the outcome is fully priced in. As we wrote in an earlier piece on this topic: the mistake most seed funds make isn’t price or pick — it’s N, the number of startups they hold.
The AngelList data supports this framing directly. Their researchers also tracked performance of investors in what they called the Access Fund — a vehicle that invested broadly across the platform’s curated deals rather than letting investors pick individual bets. Access Fund investors outperformed non-Access Fund investors in every single year cohort from 2013 through 2017. The median excess performance ranged from 1.5% to 11.1% annually depending on the vintage. In mean performance the gap was wider, driven largely by the concentration of catastrophic losses among investors who had bet heavily on just a few companies.
The AngelList researchers were honest about an alternative explanation: maybe investors who make more bets are simply better investors to begin with, and the volume is a symptom rather than a cause. It’s a fair objection. But they concluded — and it’s worth reading this carefully — that regardless of mechanism, the pattern held across every cohort and every threshold they tested. Making more investments was a replicable path to better outcomes.
Right Side Capital Management, one of the more methodical fund-level practitioners of this high volume approach, reported their Fund I at roughly 7.6x net with approximately 6.1x distributed and a 27% net IRR. 500 Startups’ Fund I (2010) showed roughly 3.8x net TVPI with a 19.5% net IRR as of Q4 2019. Peter Livingston of Unpopular Ventures, who we hosted on the Ignite Podcast (Ep. 149), has backed over 450 companies using the same philosophy — including Zepto, which returned 86x on a $25,000 check. He’s direct about what this is: not spray-and-pray, but probability-based investing in a power-law market. Neither of these is a median outcome. For context, Carta’s analysis of over 2,800 venture funds finds that a 3x net TVPI is widely considered an exemplary result, and that 2017 vintage funds at the 90th percentile cluster around 3.5x.
High-volume funds, done well, can clear that bar.
Why concentrated early-stage funds face a harder problem than they acknowledge
The case for concentration sounds compelling in the abstract: deep diligence, meaningful ownership, long-term relationships, board involvement. If you can pick the right companies, you want to own more of them.
Here is the problem. At the seed stage, picking the right companies is a task that consistently defeats even experienced investors. The companies that matter most at exit often don’t look like the obvious winners at the time of the initial check. Founders pivot. Markets shift. Breakout growth sometimes arrives three years later than expected, in a direction no one predicted.
There is also a mathematical reality to this that concentrated fund managers rarely confront directly. Arian Ghashghai, founder of Earthling VC and a guest on the Ignite Podcast (Ep. 189), put it plainly: “At pre-seed, you’re not investing in companies. You’re investing in people and problem spaces — and you hope they figure it out.” That honesty is rare in fund pitches, but it’s accurate. The uncertainty at the seed stage is structural, not a failure of diligence. AngelList’s research found that after roughly five years, a winning seed-stage investment begins drawing its return multiple from a power law where the mean is theoretically unbounded. In plain English: the cost of missing the best seed investment in any given cohort is not capped. It compounds indefinitely. Mark Suster, a well-regarded venture investor, passed on the Uber seed round. His public response when that became clear was a single word: “Aaaargh.” That reaction is proportionate. Missing one Uber doesn’t cost you the check size. It costs you everything that check would have become.
Marc Andreessen has named this structural asymmetry directly. In venture, he has argued, there are two kinds of mistakes — errors of commission, backing a company that fails, and errors of omission, passing on a company that wins. The commission error costs you 1x your check. The omission error, in a power-law market, can cost you hundreds or thousands of times that. His conclusion: the omission mistakes are much, much worse. Concentrated funds, by design, optimize against commission risk. They are built to avoid backing losers. A broad-entry strategy optimizes against omission risk instead — accepting that some investments won’t work in exchange for staying in front of the companies that might matter enormously.
Consider what this means in practice. Peter Thiel invested $500,000 in Facebook in 2004 for a roughly 10% stake. Despite dilution through later rounds, that position was worth over a billion dollars at IPO (and tax free thanks to his Roth IRA strategy). Now imagine a fund that required a minimum 15% ownership to invest — a not-uncommon ownership target among disciplined seed and lead funds. Facebook wouldn’t give it. The fund walks away. Not because they made a bad judgment. Because a rigid rule optimized against overpaying got in the way of participating. The cost wasn’t the check they didn’t write. It was everything they didn’t own.
A concentrated fund that bets on ten companies needs to be right several times over in a way that a high-volume fund simply does not. It also needs access, and access is not equally distributed. The best deals in any given cohort don’t all land in the same inbox.
A broader portfolio doesn’t eliminate the need for judgment. It extends the surface area over which judgment can operate.
The follow-on mechanism is where the math actually gets made
Getting into the game early is necessary but not sufficient. The initial check gives you information and a seat at the table. What you do next determines how much of the upside you actually capture.
Here is where a simulation study from Equationcap, a research-driven investment firm, becomes useful. They ran 10,000 Monte Carlo scenarios across a universe of early-stage VC funds and found something precise: downside risk diversifies away quickly as portfolio size grows, but beyond roughly 350 to 500 portfolio companies, adding more positions starts trimming upside without offering meaningful new protection. More companies after that point and you’re not getting safer — you’re getting average.
That finding is actually a design specification. It tells you where the floor is and where the ceiling starts. A portfolio built to operate in the range below that ceiling — broad enough to catch the outliers, not so diffuse that everything reverts to market — is exactly what a disciplined high-volume strategy should look like. The Equationcap researchers also noted that with more informed investment decisions, upside potential grows notably compared to the uninformed case. Selection still matters. It just operates on a better foundation.
This is why follow-on discipline is the second half of the system. Many experienced early-stage managers hold a portion of their fund as a reserve, deployed into the subset of companies that demonstrate real traction before their next institutional round. By that point, you’re not guessing. You’ve watched the team execute under pressure. You’ve seen whether the market responds. You’re making a decision with actual evidence rather than a thesis.
The AngelList data points in the same direction. The investors who did best weren’t just those who held many positions — they were the ones with systematic exposure built around a process. The Access Fund outperformed not because it was random, but because it was broad and curated. Volume gets you in front of the winners. Discipline determines how much of them you own when it counts.
What Team Ignite is actually doing
We don’t write checks randomly. We run a structured AI-powered evaluation process against a consistent scorecard that weights for founder quality, market dynamics, product differentiation, and early evidence of traction, back-tested against thousands of rounds. The volume in our approach comes from the breadth of deal flow we’ve built, not from dropping standards.
The deal flow that makes this possible isn’t bought through scout networks or sourced from pitch competitions. Most of it comes from inside the portfolio itself. Hundreds of active portfolio companies form a founder network that routes new opportunities to us continuously — introductions to co-founders they’ve worked with before, referrals to companies solving adjacent problems, signals on who’s building something real in spaces they know firsthand. Founders trust other founders in a way they rarely trust investors. When one of our portfolio founders says a team is worth talking to, we’re hearing something that hasn’t been packaged yet for a pitch deck.
That network also gives us something harder to quantify but more valuable over time: early signal on what’s actually working. Our founders are active builders. They see trends emerge in their own markets before those trends surface in the press or the data rooms of larger funds. A company we evaluate in month three of its existence might arrive through a connection that started because two of our founders shared a Slack channel eighteen months earlier. That kind of routing doesn’t happen through a sourcing spreadsheet.
This is why the volume in our strategy doesn’t require us to lower the bar. We evaluate over 10,000 new companies each year, and the majority of the strongest ones find us through the network rather than the other way around. We see more because we’ve built something that generates proprietary deal flow — not because we’re casting a wider net into the same pond everyone else is fishing.
When we invest, we’re building a portfolio designed to hold at least one, and ideally several, of the companies that will matter most in their sectors over the next decade. We then watch closely. Milestones get tracked. Signals get evaluated. Capital follows conviction, not habit.
The honest version of the tradeoffs
High-volume strategies typically produce better median outcomes and more consistent fund performance. They are less likely to have a catastrophic vintage where the concentrated bets all missed. They are also, at the extreme end of the right tail, somewhat less likely to produce a single legendary fund that outperforms everything.
That last point deserves a direct answer rather than a deflection. The Equationcap simulation found that portfolios beyond roughly 500 companies start reverting toward market-average returns. Infinite diversification is not a strategy — it’s an abdication of one. The reason Team Ignite runs a follow-on process with real gates is precisely because of this. We’re not trying to hold 800 positions equally. We’re trying to hold a broad initial portfolio, watch it closely, and concentrate into the companies that prove themselves. The breadth is the front end. The discipline is the back end. Neither works without the other.
As for the legendary concentrated fund: that outcome usually required two things that are nearly impossible to engineer — extraordinary access to a specific deal (Benchmark and eBay, Sequoia and Google) and a vintage where the macro cooperated. Those conditions exist. They just can’t be planned for.
What can be planned is a portfolio construction system that gives you a statistically credible path to top-decile outcomes across most vintages. Three independent research streams — observed LP portfolios from AngelList, simulation modeling from Equationcap, and a decade of fund-level data from Carta — all point toward the same design: broad enough to catch outliers, disciplined enough not to dilute them.
That is the bet we’re making. And the evidence, from real portfolios tracked over a decade, points in a direction we find compelling.
We’re not trying to be lucky once. We’re building a process that earns the right to participate in what comes next.
Sources:
“How Portfolio Size Affects Early-Stage Venture Returns” by Nigel Koh and Abraham Othman, AngelList, April 2020.
“What AngelList Data Says About Power-Law Returns in Venture Capital,” AngelList, 2019.
“Startup Growth and Venture Returns,” AngelList, 2019.
“Portfolio Construction in Venture Capital: The Impact of Portfolio Size on Risk and Returns,” Equationcap Research, December 2022.
The Horsley Bridge data was cited by Benedict Evans in his 2016 essay “In Praise of Failure.”
Team Ignite prior research:
“More Shots on Goal in a Power Law World,” Ignite Insights, December 2025;
“Can You Index Early-Stage Venture?” Ignite Insights, September 2025;
“Does Size Matter in VC?” Ignite Insights, June 2023.