Dustin Moskovitz thinks it’s often more financially rewarding to try and be an early or mid employee at a really good-looking startup than start your own. This meme has been going around a lot recently—I’ve seen lots of people make claims that there’s an unreasonable dearth of early employees relative to founders. Often there’s some implication that this is a market inefficiency because it’s glamorous to be a founder and not to be an early employee. How true is this in practice?
If it were true, it would be pretty cool, since it’s much easier to find the next Dropbox-at-100-employees than to build your own startup. But that’s why it seems implausible to me
Dustin Moskovitz spitballed that Dropbox might give 10bp of equity to their 100th hire. That seems high based on various sources (e.g. the Wealthfront startup salary tool, unfortunately now taken offline). For another thing, employee #100 at Dropbox would have joined in late 2011 or early 2012, when Dropbox was already valued at $4 billion. That would make 10bp of equity worth $4m already, which is kind of unbelievable. If that’s what they were actually paying, Moskovitz’s advice is more like “work somewhere that pays really well,” not “take advantage of a market inefficiency”—the reason you would have made money from Dropbox is more that the equity was already worth a lot than that it had high risk-adjusted returns. (Dropbox was valued at $10b in early 2014, for a return of 35% annually.)
What about other companies? As a rough proxy for being “Dropbox-level” obviously a good idea I’ll use startups that got nominated the TechCrunch “Crunchie” awards for “Best Overall Startup” in ‘08, ‘09, ‘10, ‘11, or ‘12. It’s annoying to figure out how much equity each of them actually gave employee #100, and some of them haven’t exited yet so it’s not clear what the eventual payout will be, but we can still treat someone’s equity in a company as an investment and figure out how much it grew. If the companies all give roughly the same nominal dollar value of equity to each hire (a questionable assumption to say the least!) then that will be a good proxy for how relatively lucrative it would have been to work at each one. I’ve put my work in a Google Sheet.
Of those companies, two were acquired too quickly for someone joining at that stage to see much return (Instagram and ngmoco); two were unequivocally duds so far (Zynga and Fab) and two more were probably bad ideas by the time they got nominated for a Crunchie but it’s harder to say (Tumblr, Groupon). Four more (Gilt Groupe, Hulu, Quora and Square) haven’t raised in over a year and so it’s not as clear how much of the return employees will eventually realize.
The other 9 of them had a geometric mean IRR of 58%. It’s not clear whether you should count the questionable 8 as giving essentially no return but no loss on average (which would lower the overall IRR to 27%) or what. Either way, it’s not too shabby of a return. That’s especially true if you compare it to the counterfactual of working somewhere with a high cash salary and investing it: if you work at a startup you get 4 years of vesting worth of equity at the price when you joined, which effectively gives you a big advance on the principal you can invest. Of course, there are lots of caveats, like that IRR is a pretty bad metric here since VC equity and employee equity are pretty different, or that you should take into account the present value of the option grants, not just future returns. Looking at the IRR is just a very rough first pass, nothing more.
Finally, it’s worth keeping track of which companies you might apply to if you were using this strategy today. The nominees for the 2014 Crunchies (the last set available) were GoPro, Snapchat, Stripe, Tinder and Uber.
Comments
I’ve often idly wondered how to maximize compensation. That’s not what I’m trying to maximize, so I haven’t thought about too seriously, but I’ve still wondered about it.
First, I think you have to have some baseline to measure against. IMO, if you’re considering joining a trendy startup, I think Senior Engineer at Google (or the equivalent title at FB or whatever other company you might join) is a reasonable baseline. This is a totally made up estimate, but I think that getting hired into a place like Pinterest is approximately as difficult as getting hired into Google and making it to Senior. Glassdoor lists total comp for Senior @ Google as $243k/yr. This seems a bit low to me, but Glassdoor is probably more representative than the people I’ve polled, so let’s go with that. If you think about a 10 year timeframe and naively extrapolate out, that’s $2.4m. You can try to adjust for how dev compensation is going to change, inflation, etc., but I’m just going to the simplest possible analysis here. Of course that $2.4m isn’t risk free, but it’s probably the least risk of any of the options we’re looking at, and it also has the advantage of spreading the income as evenly as possible over the time period, which gives it the best tax properties.
The next least risky thing would be to join a very late stage startup that’s about to IPO. The people I know who did this at twitter walked away with 7 figures in equity. Looking back at compensation from before the huge rise in programmer compensation, people I know who joined VMWare just before their (not particularly successful) IPO walked away with $200k-$400k in stock. Adjusting for the relative compensation then and now, that’s also probably equivalent to getting 7 figures in equity now. It’s unbelievable to me how fast developer compensation has grown, but that’s another comment. Anyway, let’s say your equity is worth a cool $1m. With four year vesting, that’s $250k/yr on top of, say, $140k/yr salary. So you end up with $390k/yr. That’s a lot better than the $243k/yr from Google. But there are some risks here. One is that the IPO just doesn’t happen and your contract includes a clause that prevents you from profiting from cashing out your equity in any way until the IPO. Another is that the company gets acquired and the terms of the deal dilute your equity to nothing. Another is that the company fails, and so on and so forth. My understanding is that, as a potential employee, a company basically can’t (legally) tell you that they’re about to have an IPO without serious implications for them, so you can never really judge the timing on this so there’s always a substantial risk of not being able to cash out your equity. Say you join one company, and for whatever reason, you can’t cash out your equity. And then you join another company four years later and it works out. So, maybe you make $140k/yr in salary for four years ($560k) and then you make the same salary for another four years ($560) plus you get $1m in equity. Ok, you still come out ahead of spending 10 years at GoogleBook, but the difference isn’t all that large.
From an expected dollar value standpoint, I suspect joining the very late stage startup works out better. If you’re earning to give, I suppose it makes sense to do that. But if you’re looking at maximizing personal satisfaction and you’re spending your money on yourself, $2.4m seems like basically an infinite amount of money (even pre-tax), at least if you have my spending habits; by going with the higher expected dollar value offer, you’re taking on extra risk for almost no extra gain.
IMO, if you try to run the numbers for late stage starts that might be, say, 2-4 years from an IPO, you end up with something similar but with a higher expected dollar value and more risk. Whether or not that’s worth it to you depends on your risk tolerance and the marginal benefit of having $4m instead of $2.4m.
Another aspect of this is that startup offers aren’t fungible. Google, FB, Amazon, etc. will usually approximately match each other’s offers and come in at around the same range. IME, and the experience of folks I’ve chatted with, this isn’t true of startups. Most startups will give you an offer that won’t match working at a big company unless the startup is wildly successful and becomes the next Google, a small handful will give you an offer that will substantially exceed what you could make at a big company if the startup enjoys moderate growth, and an even smaller pinkiefull will give you an offer that looks quite good even if the startup IPOs at its current valuation. IMO, the variation between companies is at least as important as the variation between a company at different stages.
Really interesting analysis.
I think a major problem with that IRR though is that 2009-2015 was a major bull market for equities, and technology in particular. It would be better to estimate IRR over an entire market cycle, which is going to be a lot lower.