Jun 3, 2015 · 7 minutes

When you're gonna charge $399 a year-- almost six times as much as The New Yorker -- and call yourself "the most valuable source of news about the technology industry for the world’s professionals," you should probably understand startup valuations. We don't usually offer critiques of posts from paywalled sites, but one from "The Information" yesterday was just... well, bizarre.

The post, by The Information's Tom Dotan and Peter Schulz, asked, "How Much are Workers Worth to a Unicorn?" It included a bar graph displaying the "Valuation per Employee" for five separate billion-dollar companies or "unicorns": Snapchat, Uber, Pinterest, Slack, and Zenefits.

Considering there are now close to 100 private, VC-funded "unicorns" -- more than at any other point in history -- five companies are hardly a representative sample of what Silicon Valley's Russ Hanneman would call "the three comma club." But that doesn't come close to the biggest problem with the piece: The conclusions the writers draw betray an almost total lack of knowledge of how companies are valued in today's venture capital climate.

Here's the link if you have a subscription. If not in broad strokes, this is what the data cited by the authors says: Snapchat, with a $16 billion valuation and an estimated 330 workers has the highest "Valuation per Employee" of the five startups discussed in the piece, coming in at $48.5 million. The HR software platform and insurance broker Zenefits, meanwhile, with a $4.5 billion valuation and an estimated 1000 workers, has the lowest ratio between valuation and the number of employees at $4.5 million.

The authors of the post conclude that these figures reveal "differences in the investor expectations of various companies. A labor intensive business like Uber or Zenefits has more pressures on its profit margins than an ad driven one like Snapchat or Pinterest."

So what's wrong with that assumption? Pretty much everything. For one thing valuation-to-headcount is a totally arbitrary metric, as opposed to things like market size, growth rate, competitive moat, revenue per head, gross margin per head, cost of customer acquisition, lifetime value, user engagement rates, or anything that truly measures the business. The only time anyone ever cites "valuation per headcount" is when a failed company gets acqui-hired, and the acquiring company wants to know how many engineers they are effectively "buying." 

As the tech press has written time and time again valuations indicate very little about what's going on with a business. Valuations are at best a hopeful dream of what a company could be worth one day. And valuations are only one metric in a funding round. Many VCs will tell you they'll give you any valuation you want if you let them set all the other terms. For instance, liquidation preferences can protect investors on the downside. There are plenty of hidden terms that, if revealed, would betray that investors don't always take the kind of risk they appear to take to the outside world.

In fact, the most shocking thing about Snapchat's recent round wasn't its valuation. It was the fact that investors were negotiating to buy common stock, which would give them no such protections. That's how hot Snapchat is. And valuations in hot companies are more about what price investors have to pay to get in the round than anything else.

No one thinks that Snapchat would be worth $16 billion if it went public today. Despite having as many as 200 million users, 70 percent of whom are under 25, Snapchat has only recently begun to monetize these users... by design. Ironically that gives investors more room to be optimistic on the valuation. And it behooves investors to lean toward optimism -- after all, as Henry Blodget is fond of saying, you can only lose 100 percent of your money, but you stand to lose a lot more if you fail to make a big, brave, bet on what could be the next Facebook. Even if you lose most of those bets, all it takes is one win to make you a billionaire. And this behavior helps drive up valuations far beyond what they would normally be in the public markets.

So we know valuations are, at best, wishful thinking. (At worst they are “arbitrary as fuck.") Which brings us to the other half of the ratio: Workforce.

Again, it's extremely early days for Snapchat in terms of building revenue, and therefore the company doesn't need a massive sales team to support some intricate marketing apparatus that doesn't yet exist . Like Facebook and others, you'd expect Snapchat to grow as it builds out its business. Only Google built a business model where ads seemed to sell themselves. And even Google wound up with a gargantuan staff.

But suppose we assume that the authors are correct, and that the relationship between headcount and valuation is even a tiny bit relevant -- the conclusions they reached are still flawed. Based on the data alone, it makes little sense to draw a line in the sand between "ad driven" firms like Pinterest and Snapchat and "labor intensive" firms like Uber and Zenefits. As stated above, Snapchat's "Valuation per Employee" ratio is $48.5 million while Pinterest's ratio, with an $11 billion valuation and 550 employees, is only $20 million. That's less than half of Snapchat's, making it curious that the authors would lump the two startups together as if to indicate a trend. Pinterest's ratio more closely resembles Uber's for that matter, which based on its $41 billion valuation and its estimated headcount of 3,000 employees, has a Valuation per Employee of $13.7 million. Instead, the "data" seems to prove what we just said: That Snapchat just has an outrageously gaudy valuation. 

So does that mean the authors should have lumped Pinterest and Uber together? Of course not. As a business, Pinterest has basically nothing in common with Uber beyond sharing a somewhat similar Valuation per Employee. And that's exactly why I bring it up: To underscore how completely arbitrary this metric is. 

One more point about Uber: This has to be the first article to critique its business model. The Information claims that Uber "has more pressures on its profit margins" than ad driven businesses like Snapchat. We should all have such "pressures on our profit margins." 

One on hand, it's fair to say that the ride-hailing company does have a much broader set of expenses than, say, a social media platform. Uber deals with real-life transactions that bring about various regulatory and legislative mandates that must be met, not to mention a greater likelihood for lawsuits pertaining to safety and fair treatment of riders and drivers -- all of which cuts into Uber's revenue streams.

But while as a "labor intensive" business Uber may have greater expenses than a social platform, it also has a simpler, more direct monetization route, collecting money directly from its users and at an astoundingly high rate. Despite these expenses, Uber's profits are obscene. As an illustration of this, consider that it cost Uber a one-time estimated payment of $100,000 to comply with Portland's regulations on employing immigrant cab drivers. Which I suppose is a bummer until you realize the company is on track to bring in $27 million a day in revenue by the end of 2015, $5.4 million of which it will keep in net profits. Again, a company like Uber shouldn't even be in the same graph with a company like Pinterest or Snapchat, which are just beginning to monetize. 

As for those ad driven social platforms -- which the Information improbably suggests have an easier time than Uber making profits -- they do not have the luxury of making money the old-fashioned way from users. Instead, companies like Snapchat and Pinterest are burdened with convincing marketers of a value proposition that is often difficult to measure and quantify, and that relies on subjective, still largely untested conversion metrics. (Just ask Twitter, which is further down the road and constantly under fire from investors.) The article literally argues that Uber, which is making scary godlike wads of cash from its consumers directly "has more pressures on its profit margins" than companies that don't even have a proven revenue model yet.

So are there any smart insights that a reasonable observer could draw from this data? I suppose it might be worth wondering why Snapchat is such an outlier here, especially when compared to Pinterest which is at the same early stage of revenue growth. Could it be because Snapchat is based in LA where there's a smaller talent pool of developers and other tech startup workers than in Silicon Valley where Pinterest is based? Or is it simply that Evan Spiegel is more aggressive at fundraising?

There is only one reason I can think to lump these two metrics together in a graph: They are two of the only pieces of widely available data about these intensely private companies.

[illustration by Brad Jonas]