Power Law Investing: Why the Best Seed Strategy Is Maximum Breadth

The first piece in this series argued that if you run one small business, the power law says concentrate: cut the long tail, find out what’s actually doing the work, and put more of your limited time and money behind it. Diversifying — spreading yourself across more products, more channels, more customer segments — is usually a mistake, because you can’t survive most of those bets failing.

This piece is the other side of the same law. If you’re the one deciding which startups get funded in the first place, the correct move flips completely — not because the math changes, but because the actor does. A single business owner has one shot and finite hours; a fund has many independent shots and structurally can’t tell in advance which one will pay off. Same distribution, opposite prescription.

Venture returns follow a power law. A small number of investments produce almost all of a fund’s return. Most investments return little or nothing. This isn’t a rough approximation — it’s one of the most consistently observed patterns in venture capital, and it should shape how a VC behaves. Most funds will say it does. Fewer actually build their process around what it implies.

What This Means for Seed Investing

Here’s the implication that matters most: at the seed stage, you cannot identify the outlier in advance. Not “it’s hard.” You cannot do it. The information that would let you distinguish the eventual power-law winner from the ninety-nine companies that won’t make it doesn’t exist yet — it hasn’t been generated. Product-market fit hasn’t shown up. The team hasn’t been tested by real growth. The market hasn’t revealed which version of the idea actually works.

So the question isn’t “how do I pick the winner at seed.” It’s “how do I make sure the eventual winner is in my portfolio at all.”

The Seed Stage Is a Coverage Problem, Not a Selection Problem

If you can’t identify the outlier early, the only lever you actually control is how many genuine, real shots you take. Every company you don’t fund at seed is a company that — if it turns out to be the outlier — you have zero exposure to, permanently. There’s no later round where you can buy your way back into a deal you passed on at seed, at seed prices, with seed-level ownership.

That leads to an interesting argument for a specific strategy: fund as many companies as possible at seed, each at a real, full-conviction check size — not a token stake, not a spread-thin position designed to say you were “in the deal.” The check needs to be large enough that if the company becomes the outlier, it moves the fund. But the number of companies should be as high as the fund can sustain at that check size, because at this stage, coverage is the entire game. You are not selecting the winner. You are buying the option to discover it.

This is where a lot of pattern-matching about VC gets it backwards. The image of the VC is someone with a rare gift for picking winners early. The power law argues for something less flattering and more useful: the seed-stage VC’s edge isn’t picking correctly, it’s being in enough rooms — it gets revealed later, and you’re already there when it is.

Why This Doesn’t Mean Spray-and-Pray

This is the point where the strategy gets misread as “maximize the number of logos in the portfolio, minimize the check size.” That’s a different strategy, and it defeats the purpose.

If the check is too small to matter, coverage is worthless. Owning 0.1% of the eventual outlier because the fund spread $50K checks across five hundred companies produces a rounding error, not a fund-returning outcome. The whole point of buying coverage is that when the outlier reveals itself, your stake is large enough that it matters. So “maximum breadth” has a ceiling — it’s bounded by the largest number of companies the fund can fund at a check size still worth having when one of them wins.

That ceiling is fund-specific. It depends on fund size, target ownership, and how much capital needs to be held back for what comes next.

What Happens After Seed Is a Different Problem

The breadth strategy is a seed-stage strategy. It applies precisely because, at seed, you have no signal to act on. The moment signal starts arriving — usage data, retention, revenue growth, the market’s own reaction to the product — the strategy should flip.

Post-seed, the job changes from “maximize coverage” to “concentrate behind the signal.” This is where reserves matter. A fund that spent everything achieving maximum seed coverage, with nothing held back for follow-on rounds, has bought the option to find the outlier and then can’t afford to exercise it. The companies showing real signal are exactly the ones worth doubling down on — and by definition, that’s a small subset of the portfolio, because signal, like the returns it predicts, doesn’t distribute evenly either.

So the full shape of the strategy is really two strategies in sequence, not one:

• At seed: maximize breadth. Fund as many companies as possible, at check sizes large enough to matter, because you have no way to identify the winner yet and every company you skip is permanent lost exposure if it turns out to be the one.
• After seed: concentrate on signal. Once the market starts telling you which companies are working, follow-on capital should flow disproportionately toward them — which only works if reserves were protected during the breadth phase rather than exhausted by it.

The Discipline This Actually Requires

The uncomfortable part of this strategy isn’t the seed stage. Writing a lot of checks is the easy, exciting part. The discipline is in what it requires before that — sizing the fund and the check size correctly so that maximum coverage is still affordable at a check size worth having, and holding back real reserves rather than spending them chasing more logos at seed.

Power law logic doesn’t argue for unlimited breadth. It argues for the most breadth a fund can afford while still (a) mattering when it wins, and (b) having capital left to back the winner once it’s identifiable. Funds that get this wrong tend to fail in one of two symmetric ways: too concentrated at seed, and they miss the outlier entirely because they never had exposure to it; or too spread everywhere, forever, and they have exposure to the outlier but not enough of it, and nothing left to increase it once it matters.

The strategy isn’t “invest in everything.” It’s “buy coverage while coverage is the only tool that works, then stop and concentrate the moment better tools become available.”

Putting a Number on It

The relationship above can be written as a formula. It won’t tell any fund the “right” answer — that depends on real assumptions about outcomes in a given category — but it makes the trade-off explicit instead of leaving it as a slogan.

Minimum ownership at entry, from the return-the-fund requirement:

O_min = (M × F) / E

Where:

• F = fund size
• M = target return multiple from a single outlier deal (funds often target 1x fund minimum from one deal — some target higher)
• E = realistic outlier exit value for the category/stage being invested in

This is the floor. It’s the ownership percentage a check needs to buy, given a realistic outlier outcome, for that check to actually matter if it hits.

Minimum check size, from that ownership requirement:

c_min = O_min × V_post

Where V_post is typical seed post-money valuation in the target category.

Deployable seed capital, after protecting follow-on reserves:

D = F × (1 − r)

Where r is the reserve ratio held back for concentrating on signal later — the capital set aside for the “after seed” half of the strategy above.

Maximum number of seed investments — the actual answer to how much breadth a fund can afford:

N_max = D / c_min = [F × (1 − r)] / (O_min × V_post)

This shows the breadth strategy isn’t a free variable. It’s fully determined by four inputs a fund already knows: fund size, reserve discipline, realistic exit outcomes in the category, and typical entry valuations. Plug in aggressive assumptions — a small outlier exit, a high target multiple — and N_max shrinks fast. The return-the-fund constraint is what actually caps the breadth the power law argument seems to invite.

Worked Example

The figures below are illustrative, chosen to make the math clear — not researched or sourced data on actual seed-stage outcomes, valuations, or fund construction. Plug in real numbers for a real fund.

Inputs:

• Fund size (F): $20M
• Target return from one deal (M): 1x fund
• Realistic outlier exit (E): $500M
• Typical seed post-money (V_post): $10M
• Reserve ratio (r): 50%

Calculation:

O_min = (1 × $20M) / $500M = 4%

c_min = 4% × $10M = $400K

D = $20M × (1 − 0.5) = $10M

N_max = $10M / $400K = 25 companies

Reading it: this fund can afford roughly 25 seed checks of $400K each, at 4% ownership, while holding $10M in reserve for follow-on once signal arrives. That’s the breadth the power-law argument actually supports here — not “as many as possible” in an unbounded sense, but a specific, calculable number that falls out of the fund’s own constraints.

Change any input and the number moves. Target a bigger outlier ($1B instead of $500M) and required ownership drops, so N_max rises — more breadth becomes affordable. Hold back less in reserve (r = 0.3 instead of 0.5) and N_max rises too, but at the cost of less dry powder to concentrate later — the exact trade-off the “discipline” section above describes.

Explain It Like I’m Four

Imagine a huge bin full of scratch tickets, and every ticket costs money to buy. Here’s the twist: the tickets don’t pay out a fixed prize. Instead, whatever you pay for a ticket buys you a slice of whatever that ticket turns out to be worth. Pay more, get a bigger slice. Pay less, get a smaller slice.

Almost every ticket, once scratched, turns out to be worth nothing — your slice of nothing is nothing. But a tiny few tickets turn out to be worth a fortune — say, a million dollars. You don’t know which ones in advance. It’s even possible, if you’re unlucky, that none of the tickets you can afford turn out to be one of those.

You have a fixed amount of money to spend. What’s the smart way to spend it?

Bad idea #1: Spend all your money on one ticket. You get a huge slice of that one ticket — but if it’s a loser, and most are, you’ve spent everything and gotten nothing. There might have been a winning ticket right next to it, but you never got to buy any slice of it at all.

Bad idea #2: Spend all your money buying a thousand cheap tickets, a penny each. Now you’ve probably got a slice of the winning ticket somewhere in your pile — good odds. But because you only paid a penny for that one, your slice of it is tiny. Even though it turns out to be worth a million dollars, a penny’s worth of a million-dollar ticket is still just a penny. You found the winner and it barely mattered, because you never bought a real slice of it.

The smart idea: buy as many tickets as your money allows, spending enough on each one that your slice would actually be worth something if it wins. Not one giant ticket, not a thousand pennies — the most tickets you can afford while still buying a real slice of each.

That’s what a seed investor is doing. They have a fixed amount of money (the fund). Nobody can look at a brand-new company and know it’s going to be the big winner — it’s too early, nothing has happened yet to prove it either way. So instead of betting everything on one guess, a smart investor spreads the money across a lot of small companies — buying a slice of each one — but never so small a slice that winning stops mattering. Each check has to buy a big enough slice that if that company becomes the winner, the investor’s slice is actually worth something.

And here’s the part people miss: a smart investor doesn’t spend every dollar buying tickets in the first place. They keep some money back. Because once one of the tickets starts looking special — people notice it, it starts growing — that’s the moment to buy a bigger slice of that one specific ticket. You can’t do that if you already spent everything on the first round.

One sentence version: spend your money buying real slices of as many tickets as you can afford — but hold some back, so you can buy a bigger slice of the one that’s actually winning once you know.

Where does your fund’s check size stop being coverage and start being noise?

About Me

I'm Michael Philippou, co-founder of Big Love, a plant-based ice cream business in Santa Monica that I've been running with my wife Victoria for over ten years. Before ice cream, I was a lawyer. I write about the real, unpolished lessons of running a small business — no gurus, no hype, just what's actually worked and what hasn't. Watch more on YouTube (Real Business Real Lessons) or follow along on LinkedIn.