Investors are misreading MongoDB: it’s a service provider, not a software seller
MongoDB are an enterprise IT company that went public October 19. Unlike recent consumer-tech IPOs, Mongo’s did rather well: shares rose 30% on the first day of trading. Mongo’s pitch to investors is laid out in their prospectus. The important part is here:
The company has never made a profit, but revenue from subscriptions has been growing quickly. In theory, software subscriptions are very scalable. If the business is scalable, it might not matter that there haven’t been any profits yet, because costs and revenue could be following very different curves.
If revenue is exponential but costs are linear, a company’s fortunes can go: loss, loss, loss, profit, PROFIT, PROFIT!!. Facebook and Google both made losses forever, before turning a corner and making a killing. Technology investors have learned that waiting until the company is profitable makes you much too late. Once it’s profitable the company will be buying its shares, not selling them.
This assumption about technology stocks has become so basic it’s almost never articulated. If you mention a company’s growing losses to some people in the tech industry, they look at you as though you’ve said something profoundly stupid. Of course the company is evaluated on revenue growth, not profitability. Do you know nothing?
With that in mind, it’s interesting to look at how MongoDB pitches the value it provides to its customers. The brochure and the prospectus tell very different stories about the nature of Mongo’s revenue. In the prospectus, the story is that Mongo makes its money through subscriptions. But in the brochure, Mongo lists the value of the software license as $0 — the license cost is “captured under” Software Support and Maintenance.
When Mongo are pitching their clients, the message isn’t “We have built this great software. You can use it for $35,000 a year”. The message is, “We’ve built this great software. Using it will save you tonnes of money compared to Oracle, as our software is essentially free. In order to save all this money, you’ll need to pay us $35,000 a year for our support services. We’ll throw in a gold-plated version of our software too —which you can already try for free (subject to some theoretical legal restrictions untested by any court).”
It could be that this brochure is highly misleading, and in fact Mongo spends almost nothing to provide this support. Fooling your customers is hard work though, so my hunch is that the brochure is much closer to the truth than the prospectus. Let’s have another look at the financials:
Since 2015 revenue has grown 250%, while “Cost of revenue” has grown only 60.7%. At that rate they’ll be profitable in no time, right? But hang on: operational expenses have also grown 60%. And don’t be fooled by Simpson’s paradox here: the net trend is negative. The company isn’t just losing money, its losses are growing.
From 2015 to 2017, Mongo’s Sales and Marketing expenses increased by $26m. The company is nearly 10 years old, so this wasn’t the early get-the-word-out boom of a start-up that’s just found “product market fit”. I wonder whether Mongo employ a lot of people with technical qualifications who have the job title “Customer Success”. People with this job title often sit with the sales team, but their role is very much technical support. Another job title that lets you hire engineers with your sales budget is “Developer Advocate”. I’m sure there are lots of other modern alternatives.
The problem with evaluating companies on revenue instead of profits is that Goodhart’s law ruins everything. You’re judging an actor based on a correlate of the thing you want, instead of the thing itself. As soon as your judgment starts to matter, the actor’s behaviour will change, and your metric will no longer be useful.
It’s true that if you only look for profits, you’ll probably be too late for a lot of the best opportunities. But if you don’t look for profits, you’re extremely vulnerable. The thing about profits is that they’re distinctly hard to fake. If you want to mislead me about your profitability, you probably need to tell me untrue things about the past or present. Eventually that will catch up with you. Misleading me about the scalability of your business is much less risky. You don’t have to say anything that’s necessarily false, because we’re mostly talking about the future.
The revenue-growth framing is so favourable that there are companies in all sorts of industries who see it as very, very important that they be understood as “technology” companies. WeWork don’t just rent real estate. They’re a technology company who happen to execute in the real-estate space. Tesla aren’t a car company with a shitty production process. They’re a technology company ironing out a few kinks in their hardware and supply-chain.
Everybody knows that you can’t sell dollar bills for fifty cents each and make the difference up on volume, so of course you’d never be interested in investing in a company that did that. That would be silly. But what if I told you my company sold our dollar bills for five dollars each? That sounds awesome, right? And let me tell you, sales growth is awesome because our product is awesome, and we think our customers are awesome so we send them a free bottle of wine the day after we fill their order, to make sure they tell everyone about us. We’re still fuelling the rocket ship at the moment so we’re not in the black yet, but look at that revenue growth! Awesome, right?
