# Fatal Pinches, Seed Follow-On Rates, and Estimated Marginal Dilution

Yesterday, Paul Graham posted a great article about “fatal pinches” – situations where a startup still has a decent amount of runway in the bank, but its costs are high and its revenue growth is too slow to merit another round of funding. PG suggested that startups caught in fatal pinches have 3 main options: 1) give up, 2) cut costs – which often involves laying people off, or 3) do whatever it takes to increase revenue (e.g. consulting).

I’ve talked to a number of startups in this unenviable state, and advice for how to survive it is extremely useful. However, the best advice is to try to avoid fatal pinches in the first place. A key component of that is to determine how much money one needs to raise in order to have the best chance of raising a Series A round in the future. It turns out that Tomasz Tunguz of Redpoint Ventures explored this topic a few months ago in his excellent post on seed follow-on rates. What he discovered by looking at CrunchBase data is that the odds of raising a Series A are low if you raise ~$300k, moderate if you raise ~$600k, and as good as they can be if you raise ~$900k or more.

(source: http://tomtunguz.com/seed-followon-rates/)

It’s not exactly earth-shattering to proclaim that raising a bigger seed round gives you more runway to find product-market fit and raise a Series A. However, most founders are reluctant to raise a bigger seed round because they are (understandably) paranoid about taking on too much dilution. Raising $1m at a $4m pre means you’ve sold 20% of your company, but raising $1.5m at the same valuation means you’ve sold 27.3% of your company, which is quite high if you hope to raise multiple rounds in the future. The typical mental calculation is: “I think I only need $1m, and the extra $500k wouldn’t be worth an additional 7.3% of dilution.” Most people only look at the 7.3%, without realizing that they should be looking at what I will call the *estimated marginal dilution*.

The ‘estimated marginal dilution’ is the difference between how much your capital costs now compared to how much it would cost in the future.* For example, what if you could either raise $1m at a $4m pre now and $3m at a $12m pre in 15 months, *or* you could raise $1.5m now and $2.5m in 15 months at the same valuations? In the first case, total dilution from your fundraising is 20% + 20% = 40%; in the second case, it’s 27.3% + 17.2% = 44.5%. That means that taking an additional $500k now instead of later doesn’t cost you 7.3% of your company, it costs you about 4.5%

To make the example even more interesting, what if the extra $500k runway lets you get to slightly better metrics before raising your Series A, so that instead of raising $2.5m at a $12m pre in 15 months, you could raise $2.5m at a $16m pre in 18 months? Now your dilution is 27.3% + 13.5% = 40.8% – which means there’s very little penalty for taking $500k now instead of later. If that’s the case, there’s a strong argument for taking the extra money now.

The final piece of the puzzle is figuring out what to do if you are able to raise more than you need. The truth is that you shouldn’t dramatically alter your planned spending until you’ve found product-market fit. Growing headcount prematurely is how many startups end up in the fatal pinch in the first place. Basically, if you’re raising e.g. $1m, consider raising $1.5m+, _but _run your company as if you only raised $1m until you reach product-market fit. The extra capital will give you more breathing room and decrease your chances of being stuck in a fatal pitch.

An interesting example of taking into account current and future capital costs is to look at valuations of EIR-incubated companies. The terms for those investments might be something like $5m at an $8m pre. It’s easy to look at that and think “that’s crazy, why would someone take that much dilution in their first round?” The explanation is that it’s just as good to take $5m at an $8m pre now as it is to take $1.5m at a $5m pre now, and $3.5m at a $19m pre later.