Optimal Portfolio Sizing and Fund Design for Early-Stage B2B SaaS VC Funds
Early-stage venture returns follow a highly skewed distribution – a few big winners drive most returns while the majority of investments fail or produce modest outcomes. This reality makes portfolio construction critical for pre-seed and seed funds in B2B SaaS. Below we analyze models of portfolio sizing (from simple probability math to Monte Carlo simulations and power-law models), examine how key assumptions affect outcomes, and translate these findings into practical fund design guidance (e.g. how many startups to back, check sizes, ownership targets, reserves for follow-on rounds, etc.). We also provide specific considerations for B2B SaaS-focused funds given typical sector dynamics and exit outcomes.
Portfolio Models: Binomial, Monte Carlo, and Power-Law Approaches
Early-stage investors often use quantitative models to decide how many startups to invest in. Binomial probability models are a simple starting point – they treat each investment as an independent “trial” with some probability p of a big success (an outlier that could return the fund). Using the binomial model, one can calculate the chance of getting at least one success in a portfolio of N companies. For example, if each startup has a 5% chance of becoming a “fund-returning” outlier, then a portfolio of 20 independent investments has about a 64% chance of hitting at least one (since $P(\text{no outlier})=(1-0.05)^{20}\approx36%$, so one or more ≈64%). Increasing N raises the odds of an outlier provided outcomes are independent – an assumption we examine later. While the binomial model is simplistic, it usefully illustrates why larger portfolios can reduce the risk of missing out on winners.
More sophisticated approaches use Monte Carlo simulations and power-law outcome distributions to model venture returns. Rather than assuming a fixed success probability p, these models sample the full range of possible outcomes (from total loss to huge exits) many times to see how portfolio returns distribute. Notably, venture outcomes appear to follow a power-law (heavy-tailed) distribution rather than a normal or uniform distribution. In a power-law world, most investments are mediocre or failures, but a tiny fraction achieve extreme multiples (100x+, etc.) that dominate returns. This has two important implications that the models capture:
Skewed returns: A few outliers account for a disproportionate share of returns. Industry data shows ~65% of venture investments fail to return capital, ~25% produce small returns (1–5x), and only ~10% generate returns above 5x. The extreme tail is very thin – only ~4% exceed 10x and ~0.4% exceed 50x. In other words, the odds of a “mega-winner” are very low (on the order of a few percent or less), but those rare wins can pay for dozens of losses.
Right-skewed distribution of U.S. venture returns (2004–2013) by Correlation Ventures. About 65% of financings failed to return capital (<1×), ~25% returned 1–5×, and only 0.4% achieved 50× or more. This power-law outcome pattern means a small number of big winners contribute the bulk of returns, underscoring the need for sufficient portfolio size to reliably capture those outliers.
High variance: Because outcomes are so unequal, the variance of portfolio returns is large when the number of investments is small. One fund might hit a unicorn and soar; another might miss and tank – even if both had the same strategy. This makes simulation a valuable tool: by simulating thousands of random portfolios, one can see the distribution of fund-level returns and the probability of different outcomes (losing money, breaking even, 3× fund, etc.) as a function of portfolio size. For example, Steve Crossan modeled seed-stage returns with a power-law (α≈2) and found that with just 5 investments, there was >40% chance the fund loses money, whereas at ~100–150 investments that risk fell to near zero medium.com. His simulation showed that larger portfolios consistently yielded more stable and higher multiples because the chance of capturing at least one huge winner increased with more “shots on goal.” Similarly, angel investor simulations by Kevin Dick found that a naive 20-company portfolio had a ~7% chance of losing money and often underperformed the market, whereas a portfolio on the order of 100+ investments virtually never lost money in his trials possibleinsight.com.
In practice, empirical analyses strongly echo these model findings. Crossan’s “Rational Venture Investors Should Have Bigger Portfolios” post (2018) concluded that at least ~150 investments per fund were indicated under reasonable power-law assumptions. Around the same time, VC Clint Korver examined industry data and noted that even top-tier firms (with better-than-average pick rates) need dozens of investments to reliably get an outlier; he argues that great VCs are distinguished not by clairvoyantly picking one winner but by constructing portfolios that ensure they include a winner. In Korver’s words, “VCs cannot reliably pick winners… They can, however, construct portfolios that consistently generate great returns.” Other studies (by Cambridge Associates, AngelList’s Abe Othman, etc.) likewise confirm that broad diversification – effectively “indexing” a large number of startups – tends to outperform more concentrated approaches, because an index is mathematically assured to capture all the big winners in the market. In fact, AngelList’s data science team showed that a hypothetical “invest in everything” early-stage fund would have outperformed ~75% of VC funds – simply because it didn’t miss the outliers (wow!).
In summary, a range of models (from simple probability math to complex simulations) all point to the same general advice: larger portfolios reduce risk and improve the likelihood of hitting rare, fund-making exits. However, before deciding an optimal portfolio size, it’s critical to consider the assumptions behind these models – such as whether outcomes are independent, what the true distribution of returns looks like, and how much “skill” (hit rate) the investor can assume. We will discuss these factors next.
Effects of Key Assumptions: Independence, Distribution Shape, and Hit Rates
1. Outcome Independence vs. Correlation: Most portfolio models assume each startup’s outcome is independent – like separate coin flips. In reality, venture outcomes can be correlated due to macroeconomic or sector-specific trends. For example, a downturn in enterprise software spending could hurt many B2B SaaS startups at once, or an exuberant market cycle might inflate valuations across the board. If outcomes are positively correlated, the benefit of diversification is less than the independent model predicts. A fund that backs 50 SaaS companies during a frothy market might find many of them struggle if the market turns (a form of vintage risk). Kevin Dick notes that to truly mitigate vintage risk, one might need to spread investments over multiple years – effectively having 100+ investments per year across cycles possibleinsight.com, which is usually impractical for a single fund. The practical takeaway is that if your portfolio is not diversified across different sectors or time periods, you should be cautious in interpreting model results – you may need more companies to achieve the same risk reduction that an idealized independent model would suggest. (In B2B SaaS-focused funds, this is a real concern since all investments share exposure to the enterprise software cycle.) On the positive side, some correlation can come from “rising tide” effects too – e.g. a wave of cloud adoption lifting many SaaS boats – but prudent planning assumes downturn correlations will occur at some point. In short, correlated risks mean you might require a larger portfolio (or reserves) than an independence-based model indicates, or alternatively, you diversify by investing across sub-sectors or time.
2. Power-Law vs. Uniform/Normal Distributions: Assuming a power-law distribution of outcomes (heavy tail) versus a more modest distribution greatly affects portfolio strategy. In a uniform or normal distribution (where outcomes cluster around an average and extreme values are exceedingly rare or capped), the variance of returns is low and the law of large numbers kicks in quickly – a couple dozen investments might be enough to ensure the portfolio outcome is close to the mean expectation. However, venture clearly does not follow a normal distribution – it’s the extreme outliers that generate outsized value. In a power-law scenario, the variance is so high that the law of large numbers operates much more slowly (technically, if the power-law exponent α ≤ 2, the variance is infinite and the mean itself may be undefined!). AngelList’s research estimated α ~2.3 for early-stage returns angellist.com – a heavy tail close to the boundary where the traditional “average” stops being meaningful. This means that no matter how many deals you do, your portfolio’s expected value might be driven predominantly by a single investment – the trick is you don’t know which one ahead of time. Thus, with a power-law, you need a lot more shots to statistically guarantee hitting an outlier, compared to a world of normally distributed outcomes. Many older VC portfolio models that recommended ~20–30 companies assumed a fairly tame distribution of outcomes. Newer analyses incorporate fat tails and generally conclude that those old rules of thumb severely under-diversify the portfolio. As one commentator put it, if venture outcomes are lottery-like (“extreme winners are self-similar – the top 1% of deals produce a huge share of returns, and the top 10% of that 1% produce a huge share of those returns, and so on”), then you need to “buy more tickets” to have a good chance at winning possibleinsight.com.
3. Variance in Hit Rates (Investor Skill and Strategy): Not all investors have the same probability p of finding a big winner. An established top-tier VC might convert 1 in 20 investments into an outlier, whereas an average seed fund might only see 1 in 50 or 1 in 100 reach that scale. Your assumed “hit rate” dramatically changes the required portfolio size. For instance, Korver cites Cambridge Associates data that industry-wide only ~2.5% of venture investments are true “home runs” (fund-returning level). But Horsley Bridge (an LP in many elite funds) found that the very best VCs hit outliers ~4–5% of the time. If you believe your team has superior selection skill or proprietary deal flow, you might justify a smaller portfolio because your p is higher. On the other hand, if you’re investing in an especially risky space or are less confident in picking ability, you’d assume a lower success probability and thus need more bets. To illustrate: an “average” VC with p ≈2% per deal has only about an 18% chance of seeing at least one outlier in a 10-company portfolio. A top-tier VC with p ≈4.5% would have ~37% chance with 10 companies – better, but still barely a coin flip for one big hit. Even a hypothetical superstar with p ~7% (significantly above industry norms) only gets ~50/50 odds of an outlier with 10 investments. This underscores that even excellent investors benefit from diversification – skill improves the curve (fewer deals needed for the same success probability), but it doesn’t eliminate the randomness. In Korver’s analysis, a top-tier VC targeting a 90% chance of an outlier would still need on the order of 50 investments per fund. His firm accordingly plans ~50 companies per fund, given their belief that they have ~4–5% per-deal success odds – yielding roughly 90% confidence of at least one outlier. They note that many smaller funds run concentrated portfolios (10–15 deals) and might attribute past success to skill, when statistically even random picks will hit the occasional outlier with that few shots (i.e. luck can fool us in the short run) medium.com. The lesson is to be realistic about your hit rate and not under-diversify based on over-optimistic assumptions. If you truly have an edge, a somewhat tighter portfolio may be warranted, but erring on the side of more investments is usually safer given how heavily luck factors into venture outcomes.
How Many Investments to Ensure an Outlier? (Simulations & Confidence Levels)
Using the above models, we can estimate how many startups a fund needs to invest in to achieve high confidence of getting at least one “fund-returning” outlier. This is often defined as an exit that returns 1× or more of the entire fund – e.g. one $50M outcome in a $50M fund (whether via one company’s IPO stake or acquisition proceeds). While every investment is hoped to return the fund, statistically only a small percentage will. Below are some illustrative numbers:
Average-case assumptions (p ~2% per investment): You’d need on the order of 130–150 investments to have ~95% probability of at least one fund-returning hit. Even for 90% confidence, over 110 investments might be required. This aligns with multiple studies suggesting ~100+ deals are needed for an “index-like” early-stage portfolio when success odds are low possibleinsight.com.
Top-tier assumptions (p ~4–5%): With a 4–5% chance of an outlier per deal (which is roughly what the best VC funds achieve), the model predicts needing on the order of 60–70 companies for ~95% confidence and ~50 companies for ~90% confidence of at least one big winner. Indeed, targeting ~50+ investments has been recommended by experienced investors in this category.
Optimistic/high skill (p ~7–10%): Even if one could improve hit rates to, say, 7% (superstar level) or 10%, it still takes dozens of investments. At 7%, about 40 investments give ~95% chance and ~30 investments give ~90%. At 10% (which is extremely high for seed-stage hit rates), ~30 investments yield ~95% chance and ~22 for ~90%. In other words, even an uncanny picker would want 25+ shots to be reasonably sure of a fund-maker. For most funds operating in competitive markets, a 10% hit rate is unrealistically high – so portfolios smaller than ~20 positions become very risky unless you truly believe you’re an outlier among VCs.
Another way to look at it: An “average” 20–25 company seed fund (which has historically been common) might only have on the order of a 40–60% chance of any big hit if the partners are average to slightly-above-average pickers. If they miss that one big deal, the fund is likely to just break even or worse – Seth Levine’s analysis showed that a $100M fund with 20 investments, using industry averages, would only double the fund (2.06× gross) and yield ~10% IRR, and that was with an assumption of one partial outlier; without an outlier the fund would barely return capital. Doubling the number of investments to 40 improved the odds, but even then a huge portion of the returns came from one “half-unicorn” – if that deal were missed, the 40-company fund would perform only as well as a low-risk bond portfolio. These findings reinforce that to reliably meet LP return expectations (say 3×+ gross returns), a fund almost invariably needs at least one or two home runs – and statistically, a small portfolio doesn’t guarantee you’ll get one.
In sum, larger portfolios provide higher confidence of capturing a fund-returning outlier. There is, however, a point of diminishing returns – going from 50 to 100 investments further reduces risk, but not as dramatically as going from 10 to 30. And extremely large portfolios introduce other challenges (resources to manage, smaller check sizes, etc., discussed below). The “right” number of companies is thus a trade-off between maximizing diversification benefits and maintaining the ability to invest meaningfully and support each company. With the groundwork of models laid, we now translate these insights into concrete portfolio design recommendations.
Practical Implications for Fund Design
Designing a venture fund’s portfolio involves balancing many factors: diversification vs. focus, check size and ownership, follow-on reserves, etc. The following guidelines emerge from the analysis:
Portfolio Size (Number of Startups): Conventional seed funds have ~20–30 companies, but data-driven models advocate for larger portfolios – often 50+ investments – to reliably capture outliers. As noted, Korver’s team at Ulu Ventures targets 50 companies per fund to have ~90% chance of at least one big hit. Similarly, quantitative funds like Right Side Capital and SyndicateRoom’s Access Fund aim for 100+ total investments to virtually guarantee overall positive returns. In practice, the optimal number depends on the fund’s strategy and capacity. More investments reduce risk (and tend to improve the worst-case outcomes), but too many can strain the team’s ability to source and manage deals. A realistic compromise for a seed-stage B2B SaaS fund might be on the order of 30–60 companies in the portfolio – significantly more diversified than old norms, but still achievable with a small partnership over a typical 2–3 year deployment period. This range reflects that many successful micro-VCs with 1–2 partners have moved toward larger portfolios (often using an “index plus curation” approach – a broad base of small bets, then doubling down on the most promising) rather than pure concentration. It’s important to note that large venture firms with many partners can effectively get to 50+ investments as well (e.g. 5 partners making ~2 deals/year each over 5 years = 50 deals). Problems arise when small funds apply big-fund pacing – e.g. a solo GP doing only 2–3 deals/year leading to ~10 total – that is a very concentrated portfolio and essentially bets the fund on a couple of outcomes (a “hope the coin flip goes our way” strategy). In summary, a B2B SaaS pre-seed fund should likely target on the order of dozens (not just tens) of investments. Lean toward the upper end of what your team and capital can manage, given that each additional independent bet meaningfully boosts the probability of hitting a winner.
Average Check Size and Ownership Targets: The flip side of more portfolio companies is smaller average check sizes (for a fixed fund size). However, maintaining sufficient ownership in each deal is crucial – a fund-returning exit requires that you hold a big enough stake in the winner. For pre-seed/seed B2B SaaS, typical initial ownership targets might be 5–15% in each startup. For example, investing $500k in a $5M post-money seed round yields 10% ownership. If that company eventually exits for $500M and the fund’s stake is diluted to, say, 5%, the fund’s return from this investment is $25M – which would half-return a $50M fund. To truly return the entire fund from one win, you’d either need a larger ownership or a bigger exit. Many VCs use the rule of thumb “only invest if it can potentially return the fund”, which in practice means ensuring the initial investment and ownership could be worth 1× fund in a big success scenario. For a $50M fund, a 10% ownership means the company must reach a $500M exit to yield $50M. If you only own 5%, it would need a $1B exit to hit $50M. Thus, even as you spread bets, don’t cut check sizes so small that your ownership in a winner is inconsequential. Most funds find a sweet spot where initial checks are meaningful enough (e.g. targeting 10%+ ownership at seed) and then use follow-ons to maintain a portion of that. In practice for a seed fund, this often means dividing the fund such that each initial check is 1–3% of the fund. For instance, a $50M fund deploying 50 initial checks might invest $500k (1% of fund) on average initially – perhaps a bit low for 10% ownership unless co-investing in larger rounds. Some might do ~30 initial deals at $1M each (2% of fund per initial), which gets a larger stake per company but fewer shots. There is no one-size-fits-all, but the key is to balance number of companies with meaningful stake. Extremely large portfolios (100+ companies) may force very small positions (<<1% of fund each), which can lead to insufficient ownership unless the fund accepts being a small, passive investor in rounds. Many micro-funds in fact pursue this “spray and pray” style with tiny checks; it improves diversification but at a cost of lower potential per deal. If one of your 1% positions becomes a unicorn 5 years later, you’ll do well, but perhaps not as well as a slightly more concentrated fund that owned more of that company. Thus, funds should set a target portfolio size and an average check size such that the math of “exit size × ownership = meaningful return” holds. For B2B SaaS, exits in the hundreds of millions are more common than multi-billion consumer moonshots, so aiming for ~10% ownership ensures a $300M–$500M outcome can significantly move the needle (and a unicorn outcome will be a home run for sure).
Reserves Strategy (Initial vs. Follow-on Investment): A critical element of fund design is how much capital to reserve for follow-on rounds versus invest at initial entry. Common strategies range from reserving ~50% of the fund for follow-ons (in classic Series A/B focused funds) to reserving 0% (invest everything upfront and do not participate further). Early-stage funds often reserve about 1:1 or 2:1 (two dollars initial for one dollar follow-on, or vice versa). The optimal reserves strategy depends on the value of doubling down on winners. If every investment’s outcome is independent and your edge is only at initial selection, then from a purely statistical standpoint, putting all dollars into as many companies as possible maximizes diversification. In fact, SyndicateRoom’s data-driven Access Fund does not follow-on at all – it invests in 50+ new companies per year and deploys the entire fund into first investments. Their analysis suggested that, at least for their stage, new deals on average yielded better returns than pro-rata follow-ons. The logic is that unless you have a strong information advantage to predict which portfolio companies will be huge, you’re better served by using capital to open more positions rather than doubling down. On the other hand, many experienced VCs strongly believe in backing your winners – once a company in your portfolio is clearly outperforming (hitting revenue milestones, attracting a great Series A, etc.), investing more in that company can dramatically increase the fund’s returns. This is especially true in a power-law world: if you identify a potential outlier early, concentrating additional capital into it can turn a good fund into a great fund. So should a seed fund reserve for follow-ons? The consensus approach for B2B SaaS funds is often a hybrid: reserve some capital to maintain your pro-rata ownership in the best companies through the next round or two, but not necessarily to lead big follow-on rounds yourself. For example, you might invest half your fund across 30 companies initially, and keep the other half to follow on in, say, 10–15 of those companies (the ones showing the most promise). This way, you preserve breadth but still have the ability to support winners and prevent excessive dilution. If you skip follow-ons entirely, your ownership in a breakout success will shrink a lot by exit (as later investors come in), potentially limiting how much that success returns the fund. If you over-reserve (too few initial bets), you risk missing having any winners at all. Many seed funds find a ~65–75% initial / 25–35% reserve split strikes a good balance. Ultimately, **reserves should be guided by your ability to pick winners in-progress: if you have strong conviction criteria and access to follow-on rounds, having reserves to deploy can boost returns. If not, a strategy of maximizing initial diversification (and maybe even syndicating out pro-rata rights to others) could make sense. Be intentional: model your strategy under both approaches. The extremes (all-in initial vs. heavy follow-on) each have pros/cons, and a moderate approach often yields solid diversification with some upside amplification on winners.
Pro-rata Participation and Dilution: This goes hand-in-hand with reserves. Pro-rata rights give an investor the option to maintain their ownership percentage in subsequent financing rounds. Exercising pro-rata (investing your share in the next round) prevents dilution of your stake. In fund modeling, one often assumes the fund will at least do pro-rata in the first few follow-ons for companies that are performing well. This is important in B2B SaaS because successful SaaS startups often raise multiple rounds of capital to fuel growth (Series A, B, C…). If you invest 10% at seed and then never invest again, by the time of an exit after several rounds, your stake might be, for example, 3%. If you did pro-rata in the Series A, you might keep ~7–8%. That difference can be tens of millions in a large exit. Therefore, plan your portfolio reserves such that you can exercise pro-rata on your winners at least through the key early rounds. This usually means allocating follow-on capital for maybe 1 or 2 rounds after the initial investment for, say, the top 20–30% of companies. Some funds even try to “super pro-rata” (invest more than their share) in the very best companies if possible – though often later-stage investors won’t allow much extra allocation beyond pro-rata for earlier seeds. Regardless, failing to at least maintain pro-rata in a future unicorn can be very costly to your fund’s outcome (you essentially give away a large chunk of the upside to new investors). On the flip side, insisting on participating in follow-ons of all companies is not wise either – you’ll end up throwing good money after bad on the laggards. Thus, selective pro-rata participation is key: invest your follow-on capital in the companies that are on a clear upward trajectory (revenue growth, market traction, strong lead investors in next round). This selection is another layer of skill – some funds formalize it with criteria or even use quantitative signals, while others go by intuition and experience. In any case, build your fund model assuming some fraction of initial investments get follow-ons from you, and the rest you’ll let dilute. The goal is that for the few companies that do become huge, your fund still owns enough to matter at exit. As an example guideline, a seed fund might aim to own ~8–10% at seed and still hold ~5% at exit of a big winner (through pro-rata). Hitting those numbers often produces a fund returner.
In summary, the practical design of an early-stage portfolio should marry the statistical insights (more shots = higher chance of a winner) with the operational realities (how many deals can you source and support well, and how to allocate capital per deal). There is no free lunch – you can’t just infinitely increase portfolio size without reducing check size or increasing fund size. But the evidence suggests most early-stage funds historically were under-diversified relative to the power-law nature of returns. Today’s emerging best practice is to lean toward more portfolio companies, within the constraints of maintaining investment quality and sufficient ownership. That might mean raising a slightly larger fund or adjusting strategy (e.g. more co-investments) to accommodate a bigger portfolio. It also means carefully planning reserves so that you neither dilute away your winners nor squander too much on losers.
Special Considerations for B2B SaaS Pre-Seed/Seed Funds (Team Ignite!)
Funds focused on pre-seed/seed-stage B2B SaaS have some unique dynamics to factor into portfolio strategy:
Higher Overall Success Rate (but Many Modest Exits): Data indicates that early-stage SaaS companies have a higher likelihood of achieving some exit compared to other sectors. One analysis showed an expected 78% exit rate for early-stage SaaS companies (majority via M&A, a small portion via IPO) – the highest among startup categories. This implies that outright failure rates in B2B SaaS might be lower (perhaps <25% go to zero) as many companies eventually get acquired for their product, team, or customer base. For a fund, this means you may see fewer total wipeouts and more partial returns. However, most of those exits will be acqui-hires or small acquisitions (think sub-$50M, often returning <1× or 1–3× on your investment). They help recover capital but won’t significantly contribute to fund-level returns. The power-law still applies – a few large exits will drive the fund outcome. The relatively higher “base hit” rate in SaaS is comforting (it increases the median outcome of the portfolio), but you still need the outliers for a top-performing fund. Don’t let a high percentage of companies exiting trick you into over-concentrating – those numerous small exits won’t add up to a great DPI without a big winner or two.
Typical Exit Sizes and Timelines: B2B SaaS companies that succeed can reach very large valuations, but often through steady growth and multiple funding rounds. It’s common to see successful SaaS startups exit in the few hundred million dollar range via acquisition (especially if they are capital-efficient or strategically valuable to incumbents), or to continue growing toward IPO (billion-dollar+ outcomes, but usually 7-10 years out from seed). For fund modeling, this means you might expect, say, one 10×–20× outcome, a couple of 3–5× outcomes, and many 0–1× outcomes out of a 30 company portfolio, as a plausible distribution. The “fund-returner” in SaaS might not be a 100× moonshot but perhaps a 20–30× investment that, thanks to adequate ownership and follow-on, returns a big chunk of the fund. For example, if you invested $1M in a SaaS startup at seed and owned 10%, and it exits for $300M in a strategic sale 6 years later (not a unicorn, but a great result), that’s $30M back – a 30× deal and over half of a $50M fund. If you had two of those, that’s the whole fund (plus everything else is gravy). So in SaaS, hitting a few $100M-$500M exits can often make the fund – you might not need a $5B IPO, though of course that would be fantastic. Knowing this, you should calibrate your ownership targets and support to capture those mid-nine-figure exits. If the typical exit in your portfolio is likely to be, say, $50M, then clearly a 10% stake yields $5M – nice but hardly moving a big fund. You’ll need some larger bets to pay off. Thus, maintain a portfolio size that gives you enough chances at those upper-percentile outcomes (the 90th+ percentile SaaS exits). Also, align your expectations: in B2B SaaS, unicorns ($1B+ exits) do happen (think Snowflake, Datadog, etc.), but many successful SaaS companies exit in the hundreds of millions. A power-law still holds but perhaps with a slightly “fatter middle” – a decent chunk of moderate wins. This can actually help a fund – a few 5–10× wins plus one 20× can already make a great fund. Ensure your portfolio is large enough that you’re likely to get several 5–10× outcomes and at least one 20×+. That again points to having dozens of investments, not a dozen.
Sector Correlation and Market Cycles: As mentioned, a SaaS-focused fund is exposed to enterprise software market conditions. If tech spending or SaaS valuations undergo a slump (as seen in 2020–2021 volatility), many portfolio companies could face headwinds simultaneously – slower growth, down-rounds, or tough exit markets. This correlation risk means diversify within B2B SaaS if possible – e.g. invest across different customer segments (SMB, mid-market, enterprise), different vertical domains (fintech SaaS, healthtech SaaS, dev tools, etc.), and spread out investments over your 2–3 year deployment so you’re not all in at a single valuation peak. By doing so, you increase the odds that at least some companies will thrive even if others hit a macro snag. Also, consider reserving a bit more capital for bridge rounds if a macro downturn hits – you might extend support to your stronger teams to help them weather the storm for a later exit when conditions improve. The portfolio sizing models assumed independent bets; in reality you must account for the fact that multiple portfolio companies can fail for the same external reason. A larger portfolio helps here too – with 40 companies in various sub-sectors, you’re more likely to have a couple that buck a negative trend.
Follow-on Strategy in SaaS: SaaS businesses often have metrics and KPIs that can signal future success (e.g. ARR growth, retention rates). Use this data to inform which startups to allocate your follow-on reserves to. In B2B SaaS, sometimes the winners emerge a bit later (enterprise sales cycles can be slow), but once a company hits product-market fit, the growth can be very steady and compounding. Be ready to double down on those that show consistent ARR growth and capital efficiency, as they are prime candidates for major exits or high-value acquisitions. Conversely, be cautious about over-investing in companies with only hype but not numbers – the discipline of SaaS metrics can help avoid false positives. Funds like Aligned Partners (cited by Korver as an example) follow a strategy of investing in capital-efficient SaaS companies that may exit earlier for moderate sums, but because they don’t require huge follow-on capital, the early investors don’t get diluted and can earn strong IRRs on smaller exits. This is a different model (focusing on more base hits, fewer moonshots). If your SaaS fund takes this approach, portfolio size might be a bit smaller since you’re not swinging for unicorns but aiming for a higher hit rate of moderate successes. It’s a viable niche strategy – though note that even there, outliers can surprise you. The key is aligning portfolio construction with your exit strategy: if going for a few big wins, load up the portfolio; if going for many singles/doubles, you still need diversification but maybe can concentrate slightly more on companies meeting specific lower-risk criteria.
Modeling Fund Outcomes: It can help to build a simple model for your fund with assumed distribution of outcomes (perhaps using the empirical distribution from Correlation Ventures as a starting point, or adjustments for SaaS). Simulate scenarios: e.g. “If we invest in 40 companies, and probabilities of outcomes are X% fail, Y% 1–5×, Z% 5–10×, W% 10×+, what is our expected fund return? What’s the 90th percentile outcome and 10th percentile outcome?” This can clarify trade-offs. Often you’ll find that a small increase in number of investments materially raises the worst-case (10th percentile) fund outcome by reducing chance of a total miss. For instance, Crossan’s simulation showed that at ~150 companies, none of 1000 simulated portfolios lost money – essentially the downside risk was erased by diversification. While a seed fund may not reach 150 names, even going from 20 to 40 can significantly cut the risk of capital loss and improve median outcome, according to various Monte Carlo studies. Use such analysis to justify your portfolio size to LPs: you can cite research that larger portfolios have a higher probability of outperforming and lower variance. For example, an EquityBee study found portfolios of 100+ holdings had much lower dispersion of returns and higher odds of outperformance, aligning with power-law portfolio theory equitybee.com.
Fund Size and Check Calculations: B2B SaaS pre-seed rounds might range from $500k to $2M; seed rounds maybe $2–4M. If you’re leading or co-leading these rounds, your check size might be $250k–$1M at pre-seed or $1–2M at seed to get a double-digit ownership. Ensure your fund size is appropriate for your strategy. If you plan 40 investments with average initial check $1M and also follow-ons, a $50M fund could be about right (40×$1M = $40M plus $10M for follow-ons, ignoring fees). If you only had a $20M fund, 40 investments would spread you too thin ($500k each with no follow-on, or fewer deals). Conversely, if you raised a $100M fund but still only do 40 seed investments, you’ll be writing bigger checks which could overshoot what early SaaS rounds need, or you’ll end up doing more deals (which is fine if team can handle it). So scale fund size, portfolio count, and check size in tandem. Many successful SaaS seed funds are in the $30M–$75M range for this reason – it allows a solid number of bets with meaningful initial stakes and some reserves. Very large funds (e.g. $150M seed funds) often have to either do many more deals (effectively becoming an index) or move up into Series A deals, because deploying that capital purely at seed is challenging without compromising strategy.
Realistic Hit Rate in B2B SaaS: It’s worth noting that because SaaS businesses have somewhat more predictable growth models and a high rate of smaller exits, a skilled SaaS-focused investor might have a slightly higher hit rate for “solid outcomes” (e.g. maybe 15–20% of portfolio companies get acquired for >1× return). But the rate of truly spectacular outcomes (10–20×+) is still low. The Cambridge Associates figure of ~2.5% “blockbuster” rate likely includes many SaaS deals. Unless you have reason to believe your proprietary deal flow (say, all the top SaaS startups from a certain accelerator) gives you a much higher success probability, plan on the low single-digit percentages for big wins. That means if you do 30 deals, you’d be lucky to have one big winner; 50 deals gives you maybe one or two; 70–100 deals almost assures one, etc. The math doesn’t change by sector – it’s just that in SaaS your “big win” might be a bit more attainable (say a $1B IPO is not science fiction for an enterprise startup with strong metrics, whereas in some sectors unicorns are extremely rare). This can be motivating: you know if you assemble a broad portfolio of SaaS startups, odds are good that a few will ride the secular trend of cloud software and become very valuable. But you likely need that breadth to hit the right one.
Support and Value-Add vs. Portfolio Breadth: One pushback on large portfolios is whether a small team can meaningfully support that many companies. This is a valid concern – part of a VC’s value proposition is the help and attention they provide to founders. If you invest in 50+ companies with 1–2 partners, clearly you can’t sit on 50 boards or be deeply involved with each. Some funds address this by adjusting their involvement model: Korver mentions that at Ulu they roll off boards after Series A and transition to a lighter support role so they can free up bandwidth for new investments. Others take an approach of being available on-demand rather than scheduling intensive oversight with each startup. There’s also the philosophy that at pre-seed, the best help you can give is often simply cash and a few key connections, and the founders will largely drive outcomes. So a trade-off exists between portfolio size and depth of engagement. If your fund’s strategy is to be very hands-on (incubation style or heavy coaching), you may need to keep the portfolio smaller. If your strategy is more about picking and assembling a large option set (like an indexer), you accept a more scalable support model. B2B SaaS investors often pride themselves on domain expertise and networks (e.g. helping companies land pilot customers, make key hires in sales, etc.), so you should strike a balance where you can still deliver that value. Some firms solve this by leveraging venture partners or advisors to extend their capacity across more companies.
To conclude, for a pre-seed/seed B2B SaaS fund, the evidence strongly supports holding a larger, diversified portfolio – likely on the order of 40–50+ companies (or even 100+) – to maximize the chance of a fund-making outcome. This should be coupled with thoughtful fund construction: right-sized checks to secure meaningful ownership, a reserve allocation to back the breakout SaaS stars, and a strategy to manage a broad portfolio (through team (ignite) structure or support network). Such a design acknowledges the power-law nature of venture returns while also tailoring to the SaaS context (where exits are frequent but often require significant scale to deliver large multiples). By comparing models and empirical data, we see that the “optimal” early-stage SaaS fund skews larger in portfolio size than traditional wisdom held, mitigates risk via diversification, and still positions itself to capture the upside of those rare outlier successes. As one VC investor quipped, “VCs need unicorns just to survive”, and thus the goal of portfolio strategy is to maximize the probability of finding (and funding sufficiently) at least one unicorn or equivalent big winner. By following the above principles, a fund can tilt the odds in its favor in the high-risk, high-reward game of early-stage B2B SaaS investing.
Sources:
Crossan, S. “Modelling suggests rational venture investors should have bigger portfolios.” Medium (2018) – Power-law simulation indicating ~150 investments for stable returns medium.commedium.com.
Korver, C. “Picking winners is a myth, but the PowerLaw is not.” Medium/Ulu Ventures (2018) – Analysis of outlier probabilities (2.5% industry hit rate vs 4.5% top-tier) and argument for ~50+ portfolio size medium.commedium.com.
Dick, K. “Simulating Angel Investment: Kevin’s Remix.” Possible Insight (2010) – Monte Carlo simulation recommending 100–150 investments for seed funds to achieve diversified returns possibleinsight.com.
Levine, S. “Venture Outcomes are Even More Skewed Than You Think.” SethLev ine.com (2014) – Correlation Ventures data on outcome distribution (65% <1×; 4% >10×)sethlevine.com and implications for portfolio returns sethlevine.com.
Othman, A. “What AngelList Data Says About Power-Law Returns in VC.” AngelList Blog (2020) – Empirical confirmation of extreme power-law (α≈2.3) in early-stage returns and benefits of broad indexing angellist.comangellist.com.
Others: Cambridge Associates report on “Breaking the Concentration Curse” (2015) medium.com; Pitchbook data on SaaS exit rates wscouncil.in; Jason Lemkin (SaaStr) on fund math requiring $2B exits per $50M fund saastr.com; SyndicateRoom Access Fund whitepaper (2022) on diversification and no-follow-on strategy files.syndicateroom.com; EquityBee VPO model (2023) on 100+ portfolio diversification benefits equitybee.com; and additional industry commentary medium.commedium.com.