Cloudera filed an S-1 last Friday, giving outside investors their first look into the financials of this high-flying unicorn. Using my framework for valuing SaaS companies, the bankers have their work cut out for them as the analysis suggests the company will almost certainly IPO at a significant discount to the $4bn+ post-money valuation it received in 2014.
Key takeaways from the filing:
1) Cloudera is big and growing fast- the company generated $200m of subscription revenue in 2016, growing 68% Y/y, which is actually faster than it grew in 2015.
2) Gross margins expanded nicely in both the subscription and services segments- from 55% overall in 2015 to 67% in 2016.
3) Growth has slowed lately, from 83% Y/y subscription growth in Q2 of 2016 to 53% growth in Q4. Analysts will be hard-pressed not to forecast further slowing in 2017 unless the company presents good reasons why this oughtn't happen in the IPO roadshow.
4) Cloudera is deeply lossmaking, with a -72% GAAP net income margin and a -47% free cash flow (FCF) margin in 2016.
5) Losses improved since 2015, but only modestly on the FCF side, which is understandable given that growth didn't slow. Net income margins expanded more, but this seems largely due to some sort of stock award in Q2 of 2015. Comparing Q4 2016 with Q4 of 2015 shows virtually no margin improvement.
Overall, I don't expect investors to treat Cloudera kindly relative to its 2014 valuation. This is not due to the performance of the company: it will generate ~$350m of revenue this year less than a decade after its founding, an impressive feat. Instead, it is due to the very, very stretched valuation it received in 2014. Though it doesn't have best in class growth efficiency like some recent SaaS IPOs and is unlikely to garner a super-premium valuation like MuleSoft and Okta, Jimmy McMillan explains the primary cause of the pending down round best:
The meme is plural because Cloudera is far from alone in this- dozens of unicorns are likely in a similar position. It just happens to be one of the first of them to file an S-1.
Benchmarking and valuation thoughts:
Here's a look at how the company's growth rate/losses measure up to peers:
The positioning in the top left of the plot puts Cloudera in the same camp as Okta: high growth, high losses. However, a few factors point to a story a bit weaker than the plot suggests:
1) Gross profits were depressed in 2015 by what looks like a one-time stock based compensation expense, boosting 2016 growth by about 10%.
2) Cloudera's topline growth has slowed through 2016, with Q4 gross profit growth of ~70%. Most SaaS companies see growth gradually slowing over time, by Cloudera's pace of deceleration is faster than what we typically see.
The valuation comparison charts have a much clearer message. However you slice it, Cloudera seems unlikely to achieve anything close to a $4bn valuation:
These charts were built using a $4bn market cap and a $3.8bn enterprise value, and it is hard to see how investors will be comfortable buying Cloudera shares at a valuation over 10x 2017 gross profit based on its financial metrics. That would imply an enterprise value of $2.5bn and (depending on the size/composition of the eventual raise) a market capitalization near $3bn- an impressive achievement by any measure, but a disappointment to investors who valued the company over $4bn three years ago.
Though the market is full of surprises and IPO roadshows matter, I would be inclined to expect a valuation in the ~6x gross profit range, or an enterprise value of about $1.5bn, given the slowing growth and the fact that on a revenue growth basis (chart at the left) Cloudera looks worse (the market often focuses on revenue growth and multiples, though I disagree with that).
Cloudera will also be weighed down by the struggles of its closest public comparable, Hortonworks (HDP), another Hadoop distributor that trades at just over 2x EV/2017 sales (I haven't modeled HDP so I don't have a gross profit estimate for it). Cloudera deserves a premium to Hortonworks, but even 3x EV/2017 sales implies just a $1.05bn enterprise value.
Overall the picture here is unusually unclear. It is clear that Cloudera will ultimately price significantly below a $4bn enterprise value, but how much lower is tough to gauge. I don't envy the investment bankers on the deal managing various competing interests/notions of value- but if I were them I would think of an initial range that values Cloudera at an enterprise value between $1-1.2bn, to ensure the success of the deal with an option to raise the range should the roadshow go well. That's a tough pill to swallow for a company that has raised almost that much money, but a better outcome than an IPO set to struggle no matter how the company performs in the next year.
What to do when the last valuation was simply too high:
I'll end with some words of advice for management teams at any stage in positions like this where a valuation from a prior round is simply unachievable despite strong execution and objective success from a business perspective. Valuation is important for all the obvious reasons, but it is just a number, and while it is generally a good idea to avoid possessing a valuation too far above the true intrinsic value of your company, once you're there the right answer is to ignore the past and raise money at what investors think the company is worth, without trying to wrangle a higher valuation that might not last to save face. In other words, take the hit up front and get to a point where intrinsic value and valuation are aligned so that further business progress isn't paired with further valuation disappointment.
The investors you want (both VCs and in public markets) will be honest with you about what they think your company is worth and why they think so, not necessarily cheerleaders willing to give you the highest valuation possible at any one time. If you feel that your investors are more bullish about your company than you are yourself and are valuing it that way, that's a red-flag, not a green one. This is especially true as companies enter a phase where their valuation is tracked carefully whether in databases like Crunchbase or down to the minute in public markets. It is better to be in a position where the chart can be up and to the right with good execution than one where it is doomed to struggle against a stretched initial valuation brought about by artificially high expectations from a bygone era.