(There’s No) Accounting for the Subscription Economy

You may remember the subscription economy from previous posts. It’s one way to make sense of cloud computing and the many new and very different ways of doing business on the Internet. We’re most familiar with Software as a Service and how different it is from conventional licenses; so familiar, in fact, that I don’t need to describe it for you here.

But subscriptions as a way of doing business are just about everywhere; they’re not just in tech anymore. For instance, if you want, you can get your clothing as a subscription, and not only that but men (who as a group are notoriously lazy shoppers) have sites dedicated just to them. You know the trend has arrived when something like men’s clothing is available as a subscription.

Nonetheless, we’ve more or less glossed over everything below the waterline in this new approach to business. It’s taken over 10 years to get the idea of the subscription economy into our noggins, but we’ve barely started internalizing what it takes to support it and report on it as a business.

Unusual Revenue Situation

This all came into sharp focus for me last week when I reviewed Salesforce.com’s Q4 and annual earnings call with Tien Tzou, CEO of Zuora, a company that specializes in what’s below the subscriptions waterline. Tzuo is also an alumnus of Salesforce, having been its CMO and chief strategy officer before starting Zuora.

As you know, subscriptions operate through customer payments on a periodic basis. The industry became known by its per-seat per-month pricing, but that doesn’t happen much these days because monthly billing got to be a challenge with big deals. Today customers sign contracts for a fixed length of time and vendors invoice periodically. A typical example might be a three-year contract with annual or quarterly billing. Here’s where it gets interesting.

The financial analysts and other Wall Street types — whom I have absolutely nothing in common with — are very accustomed to companies selling products rather than subscriptions and collecting the money net 30 or whatever and moving on to the next opportunity. Subscriptions have a mixed bag of revenue recognition ideas that challenge the status quo (which has very well-defined ways of recognizing revenue) significantly. Product companies don’t have much when it comes to reliably forecasting future revenue streams, but subscription companies are just bristling with information. Take the Salesforce revenue numbers from last week’s earnings call as an example, and here is where I am indebted to Tzuo for his insights:

• Quarterly revenue of US$632 Million, up 38 percent year-over-year
• Full-year revenue of $2.27 billion, up 37 percent year-over-year
• Deferred revenue of $1.38 billion, up 48 percent year-over-year
• Unbilled deferred revenue of $2.2 billion, up from $1.5 billion year-over-year

If you are reading this (thank you very much), you have at least an intuitive understanding of revenue, but deferred and deferred and unbilled revenue deserve explanation, because who really cares about unbilled deferred revenue — isn’t that complete vapor?

As Tien Tzuo said to me, think of it this way: You do a deal with a company in which you agree to supply your service for three years for $36,000, or $1,000 per month, and you agree to invoice once annually, in advance, for $12,000. At the very beginning, then, you have $24,000 in unbilled deferred revenue and, since you bill in advance, you also have $11,000 in deferred revenue and $1,000 in real live revenue which you can recognize.

This $1,000 is also known as “MRR,” or “monthly recurring revenue.” Theoretically, if you add up all the MRRs on the books you can get very close to the forecast for the quarter. But there’s also an upside possibility that you’ll sell something else. If you do and you invoice for it, you’ll add to that pile of money. Unfortunately, there is also a possibility that some of your MRR will go away either because the customer quit or because they didn’t renew or whatever. We know this as “churn,” so you really need to discount the MRR by the churn rate to get a better sense. Life would be simpler if we could all agree on using a metric called the “annual recurring revenue,” but curiously, ARR doesn’t exist yet.

The Lump Factor

So, all this has the potential to drive Wall Street types nuts. They’re good with the $1,000 in MRR, and they can tolerate the $11,000 in deferred revenue because it’s in hand, and the $24,000 in unbilled deferred revenue is sort of OK (but not really) because there’s a contract in place that defines the annual billings. But this does have one effect that many financial types like — it smoothes out the revenue stream for months in advance. Bookings might fluctuate, but the monthly revenue stream should be rather predictable.

Nevertheless, it’s bookings that have recently made some people skittish. Sales has always been a lumpy affair. In some months many deals get booked, and in other months not so much. Early on, the software industry trained its customers to wait until the end of the quarter to make purchases because that’s when they had leverage. Finance guys didn’t like this, but they got used to it.

Today, the quarterly incentive is largely gone due to monthly recurring revenue, but people still obsess over bookings. What if bookings go down for a few months? The logical answer is that future revenue would eventually feel it, but it’s equally true that bookings could recover before real revenue took a hit, in which case the fluctuation in bookings would not be seen. Call it seasonality.

Let’s summarize all this. Salesforce has $1.38 billion in deferred revenue, which I presume will be realized in the next 12 months. During that time they are advising us that the company will have revenues of between $2.92 and $2.95 billion. This means that they have about 47 percent of next year in the bank. They also have $2.2 billion under contract to be invoiced (unbilled and deferred), and some of this invoicing will be done at some point beyond the next year. In the last quarter, Salesforce had $632 million in revenue, which grew at 38 percent year over year. At some point in the next twelve months, Salesforce could have a quarter in which it books revenue of $750 million, which would give it a forward-looking run rate of $3 billion.

It’s still an uphill battle explaining revenue recognition and the difference between conventional companies and subscription companies, but at least there’s a lot of black ink to do it with.