Scenario Engine and Follow-On Strategy
What your follow-on rate says about your firm
Two funds can set aside the same reserve and use it very differently. Follow-on participation and check concentration show the pattern. The next step is understanding the decisions and constraints behind it.
Look beyond the reserve percentage
Two funds can hold identical reserve percentages and behave nothing alike. One puts follow-on money into most of its portfolio in small amounts. The other puts it into a handful of companies in large amounts. Same reserve, opposite strategies, and the reserve figure discloses neither.
Two measures help distinguish those approaches: the follow-on rate, or share of companies receiving more capital, and the concentration of that spending. They describe the pattern. Understanding the reasons takes a closer look at company needs, access, terms and the fund’s decisions.
What two samples show
Across the Blue Future Partners survey of emerging managers, the number of companies receiving a follow-on drops by roughly half at each successive round, while the median follow-on cheque rises about 78 percent above the previous one. The characteristic shape is a narrowing funnel with a rising cheque: fewer companies, more money each.
Some of that narrowing is not a choice. CB Insights tracked more than 1,100 US companies that raised seed rounds between 2008 and 2010 and found 48 percent reaching a Series A, around 30 percent a Series B, and 15 percent a Series C, with nearly 67 percent stalling somewhere without exiting or raising again. A fund cannot follow into a round that never happens. Those historical cohorts provide context, not an estimate of how much attrition explains any particular fund’s follow-on rate.
To understand a fund’s own pattern, first identify which companies raised again and where participation was available. Then examine which opportunities the fund chose, declined or could not fund. A portfolio- wide rate alone cannot separate those explanations.
Two coherent positions
Charles Hudson and Eric Paley offer different practitioner perspectives on how a seed fund should use follow-on capital.
Charles Hudson took the empirical route. His firm classified every follow-on cheque it had ever written as either offence, meaning they wanted more of a company they believed in, or defence, meaning something else, usually a bridge for a company that needed one. The audit found both categories were called correctly about 75 to 80 percent of the time, in that retrospective assessment. His conclusion was not to do more of both but to separate them sharply: be aggressive on offence, and write the smallest cheque possible, including zero, on defence. His firm reserved about a quarter of the fund and kept it for roughly the top fifth of the portfolio.
Eric Paley took the structural route to the opposite answer. His position is that venture funds are made on the first cheque and destroyed on the follow-on cheques. The argument is about cost basis: a seed fund that reserves heavily ends up with a weighted average cost basis at Series B prices while still describing itself as a seed investor, which forfeits the one structural advantage seed investing has. The first cheque, he argues, is the highest-returning cheque any fund writes, and capital moved out of first cheques and into later ones is capital moved to a worse price.
These arguments emphasize different things: Hudson’s review of decisions at his firm, and Paley’s concern about the price paid for later exposure. Neither establishes that another manager’s approach reveals how well that manager can select future winners.
Questions to ask of the pattern
Read the rate and check concentration together, then ask what produced them. Strategy matters, but so do financing needs, participation rights, round access and available capital.
Small follow-ons across many companies may preserve options. Larger follow-ons into fewer companies may concentrate exposure. Either pattern needs an explanation, but neither is a direct measure of confidence or investment skill.
Broad participation with large checks can use reserves quickly, so compare the plan with the available budget. For that capacity question, see the guide to running short of follow-on capital. A large unused reserve also deserves review: it may reflect fewer suitable opportunities, changed timing or an initial budget that no longer fits. The implications depend on the fund’s terms and remaining investment window.
Where the model reads it
The platform records follow-on rounds against each position, so a firm's realized follow-on rate and the shape of its cheques are visible in its own data rather than in a survey. That record is what turns this from an abstraction into something a General Partner can check: how many companies actually received a second cheque, how large those cheques were relative to the first, and how that compares to what the fund said it would do.
The model does not score the answer. There is no correct follow-on rate, and a firm at either pole is doing something defensible. Use the recorded rounds and reserve deployment to compare what happened with the declared plan. A difference is a starting point for review, not proof that the decisions were wrong.
Sources
- An LP take on VC portfolio constructionRodrigo Ferreira, Blue Future Partners, published on OpenVCThe shape of follow-on behaviour across sixty-plus emerging managers: how participation narrows by round and how cheque sizes move.
- The Venture Capital FunnelCB InsightsRound-by-round survival for more than 1,100 US seed companies from the 2008 to 2010 cohorts, tracked through 2018.
- Lessons from 600+ Investments, Founder Profiles that Win, Reserve Strategies that Drive Returns, and the Hidden Potential in Consumer StandoutsCharles Hudson on The Full Ratchet, episode 488The offence and defence audit of every follow-on cheque the firm had written.
- Dispelling Conventional Wisdom in VC, Part 2: Should Seed Investors Follow-on?Eric Paley on The Full Ratchet, episode 136The case that heavy follow-on drags a seed fund's cost basis to Series B levels.
