It is no secret that we've had a scarcity of exits lately, and as a result the makings for an epic run of IPOs in the next few years. Well over a hundred tech companies have raised capital at $1bn+ valuations, and many of these did so with promises to late stage investors of imminent liquidity in "pre-IPO" rounds. There are some legitimate reasons why "pre" has stretched into years: shielded from public eyes and quarterly results calls, companies are able to make bold strategy shifts and aggressive investments that would give public markets investors whiplash but may be necessary in today's fast-changing landscape. In many cases, however, there are other culprits: friends in the banking world have told me that many companies are holding off in the hopes of being valued on "2018 numbers" in order to avoid an IPO down-round after raising at an overly optimistic valuation. More worrisome, the data suggests a large proportion of unicorns have seen their growth slow precipitously, raising the prospect that some may never be able to IPO at anywhere close to their high-watermark valuations.
A side effect of unicorns' sheepishness when it comes to going public is that the only thing rarer than unicorns is their financial statements. This lack of transparency leaves us guessing, reliant on valuation changes at mutual funds that hold unicorn shares to get a sense of how they're faring.
There is one data source that can give us clues into how they're doing, though. Thanks to LinkedIn, we can see employment metrics going back two years for almost all of them. Pairing this data with equivalent numbers for a set of public companies can lend some tentative insights into how they're doing, both as a group and individually.
Here's the data, which covers about 100 unicorns with some exclusions for various reasons. This histogram should be pretty self explanatory, with companies sorted into employment growth buckets and color coded by the year of their entry into the unicorn ranks:
There are a few key takeaways from looking at the data this way:
1) Older unicorns (in red, oranges and coral) tend to be growing hiring more slowly, with notable exceptions like Slack, Magic Leap and CloudFlare.
2) Unsurprisingly, there is some stickiness around the 0% mark that interrupts what might otherwise be a more normal distribution- companies avoid shrinking headcount at all costs.
3) A surprising amount of unicorns are growth headcount slowly (<10% per year) or shrinking outright. Almost 40% fall into one of those two categories.
4) There's a rightward skew of hypergrowth companies, including a few which grew too fast to fit on our chart (hat tip to Cylance and Jetsmart)
5) The group in the chart employs 116,000 people today, up from 73,000 two years ago and 100,000 at this time last year.
What does this tell us about unicorn valuations? Unfortunately, there are huge limitations in employment data alone- it tells us nothing about revenue or cost per employee and so we have no direct insight into how fast unicorns are growing or how profitable they are. However, we can use public companies as a comparable dataset to try to draw from tentative conclusions.
First, there's a decent association between revenue growth and employment growth for public companies:
Comparing this to the histogram, we can make some observations:
1) Most public high-growth tech companies I've modeled are growing headcount faster than 10%.
2) On average, these companies are growing revenue a bit more than twice as fast as headcount.
3) Comparing the two sets, unicorns have a higher variance- more unicorns are shrinking, and more are growing headcount faster than 25%.
If employment growth correlates to revenue, what about valuation multiples? Its a bit more tenuous, since profitability is a big driver of valuation, but there is still at least some relationship:
Here, we can see quite clearly that the vast majority of companies with the sort of premium valuation that is likely to make most unicorn investors happy are growing headcount faster than 10%. Of those growing more slowly with big multiples (highlighted), all share a rare trait: GAAP profitability.
There are too many other variables at play to make firm predictions about what any given unicorn is worth, of course, most importantly current profitability- we'll have to wait for S-1s to get a better handle (my database is ready for more tickers!). Still, it is safe to assume that few unicorns planned to shrink or slow headcount growth to a crawl when they made their pitches to investors, and I'd be surprised if many of the companies on the list are GAAP profitable anytime soon. Put together, those point to the potential for some unhappy exits in the years ahead.
Finally, here is what the histogram looked like just a year ago. Times have changed!