Interestinthings Law, Startups, Music; maybe in that order

The Curious Case of the $150M Series A

I was reading through the feed for Yokum Taku's excellent Startup Company Lawyer blog, and came across a really interesting post on convertible debt financings with a price cap. Convertible debt is often the easiest way to do a seed/angel stage financing, because it avoids some of the complexity of selling actual equity in your startup; most importantly setting a valuation. Instead of selling shares today, you take the investors' money as debt that can convert into Series A shares, at some discount to that Series A price. So the seed investors punt the valuation decision to the next round and get compensated for their extra risk by getting a lower price when that round comes around. That's why it can be helpful to think about the seed round as a "pre-A" financing (later convertible debt investments between funding rounds that work the same way are called "bridges"). Nevertheless, Yokum says standing alone, they represent a bad deal for the angel investor, which I found a little surprising, because to my knowledge, they have been done like this for quite some time. It may not be standard for every seed deal, but if it were actually crappy, I would expect to barely see it all.

And herein is the issue. Yokum's example to show a poor outcome for the investor presumes a $500K angel investment and then a $50M Series A investment at a pre-money valuation of $100M. This means that the angel will convert into a 0.4% ownership stake in the company. Now obviously that is not what the angel had in mind; if a startup is doing that well, it's supposed to be the home run return that makes up for all their other misses, so they want to have a much bigger piece of the upside. But that's just it: that startup is doing absurdly well, maybe even inconceivably well. It means they've taken $500K (and whatever other assets they had lying around) and created $100M in value.

So is this a crazy story, at least for web startups? It's not news that it's gotten much much cheaper to do a web startup, but I think this level of lean must be mostly accounted for by a self-sustaining growth model, where new users each pay for themselves. Now that kind of model is most simply implemented by actually asking users to pay for your product, but I think the more interesting question is if it can be implemented while still making the user experience free of charge, as most web startups do, and I think the answer may be yes. Working backwards through Yokum's example, if you're worth $100M, you probably have crossed a significant inflection point in your user growth, but you had to somehow get there with a total capital raise of around $500K. That's just not that much cash to burn, and you have to pay your people something, leaving not that much to spend on actual service provision, and even less on capital investments like servers. So you had to be able to scale incrementally, and that means this is really a story about the efficient conversion of incremental cloud-computing resources like Amazon EC2 and some quality incremental monetization methods, such that the resources needed to serve each user can be purchased with the revenue from the ads shown to them (there's certainly other ways to monetize, but ads are a good example because they can theoretically be as incremental as the resources).

Framed this way, the story becomes a lot more plausible to me, and it's an interesting example of how cloud computing may actually have affected the kinds of investment deals that get done, not just who gets funded by those deals. Essentially, the traditional venture chain of multiple funding rounds is getting collapsed in some instances, such that you still need a seed round to start, but you may not need anything more until you need a lot, in order to get truly massive scale. The idea of convertible debt then breaks down, because you can't necessarily expect the next round to happen before the value of the company changes too much, so you can't rely on that round to set your share price. If you want to understand how the addition of a price cap solves this issue, you should really go read Yokum's post.