Calculated Billings: A Murky Crystal Ball for Predicting Future Performance
It is interesting that the financial markets that appear to value subscription revenue much more generously than perpetual license / maintenance revenue are now turning to calculated billings, a far more volatile metric, to predict the future performance of subscription businesses.
For those who may not be familiar with this metric, calculated billings is used by investors as a way to evaluate subscription businesses that do not directly disclose billings as part of their financial reporting. Calculated billings is defined as revenue plus the sequential change in total deferred revenue as presented on the balance sheet.
The rationale for using such metric is that if calculated billings grows faster than revenue in a given period, it would suggest that future revenue growth would eventually increase to reach the billings growth number. Conversely, if calculated billings are growing more slowly than revenue, it would suggest that future revenue growth will decrease.
Seem simple enough? Well, it’s not.
The truth is sometimes slower calculated billings growth might actually be positive for the business and faster growth less than ideal. Here are a few reasons why:
- Billings are impacted by pre-billed subscription durations, an element that is often ignored by investors: Most software subscription companies have a policy for customers that mandates the first year of subscription be pre-paid. It can also happen that a customer may want to lock in a price for a long duration and pre-pay two or three years of subscription. In most cases, if the customer pre-pays multi-year subscriptions, they will also receive a discount for that upfront payment. Reaching a billings target (such as a Calculated Billings consensus) by increasing the subscription duration may not necessarily be good for the business as it may be achieved through higher discounting (and thereby lower future revenue). Additionally, if billings growth is primarily due to increased pre-billed duration, it will not necessarily imply that future revenue will grow at the same rate as billings. To take a simple example, if calculated billings grow by 100% because pre-billed duration has doubled, this would have no impact on revenue growth as these billings will now be recognized as revenue over a duration that is twice as long.
- Co-terming of subscription agreements will most likely lower billings but benefit both customers and the company: If a subscription vendor intends to do repeat upsells or cross-sells to its existing customers, it is generally a good idea to align the end-dates of new subscription agreements so that new and existing agreement all renew on the same date. This makes future renewals a win-win for both the vendor and the customer as the renewal cycle is more efficient and easier to manage. However, this may result in stub period subscription agreements that are shorter than one year and thereby impact billings and short-term deferred revenue.
- Calculated billings will not differentiate renewal billings from new business billings:Imagine that a subscription company launches a campaign to its existing customers, where customers can renew their one-year subscription for two years and benefit from a special one-time discount of 20% if the two-year subscription is fully pre-paid. Also, imagine that this subscription company does not book any new business either from existing customers or new ones. The deferred revenue will increase substantially due to the increase in renewal billings while in fact, the future revenue of the company will actually decrease as the renewed revenue is now discounted by the special 20% discount.
- Renewal billings timing: If average pre-billed subscription duration changes or customers are either late or early in their subscription renewals, it can have a substantial impact on calculated billings without potentially having any impact on revenue. If a customer renews a subscription early, the timing of billing will be impacted, but the revenue will still be recognized over the contracted subscription duration period.
- Professional Services will impact the overall billing number: Often, software vendors do not detail their deferred revenue by either subscription revenue or professional services revenue. Fluctuations in professional services activity will, therefore, impact billings and result in misleading expected total revenue growth.
So, if calculated billings is an unreliable indicator of revenue, what other metric could be used to forecast future performance? In my opinion, there are a few different options:
- New Annual Contract Value (ACV) Bookings: a very rare number of companies disclose this information and it is likely one of the best metrics to predict future revenue growth. The challenge with disclosing ACV bookings is its lack of common definition across software vendors. For example, Company A may annualize a 5-month subscription while Company B may not. Company A may recognize a booking at time of contract signature which may not necessarily be at the start of the subscription period, while Company B may recognize that booking only at the start of the subscription period. The lack of a common standard and definition may lead to erroneous conclusions.
- Using Short-Term (ST) deferred revenue instead of total deferred revenue – Using short term deferred revenue instead of total deferred revenue to calculate “Calculated Billings” will yield a result that will be very different than normal calculated billings; it is a proxy for an “annual calculated billings” with an objective of eliminating the impact of duration on billings. However, using short-term deferred will not:
- Remove the impact of the timing of renewal billings (i.e. early or late renewals relative to the subscription expiration date),
- Normalize for the weight of renewal billings new business billings. Short Term deferred revenue is generated from three sources: 1. the transfer of Long Term deferred revenue into short term, 2. new business that is booked and which will be recognized over the short term (i.e. next twelve months) and 3. renewal bookings that will be recognized over the short term. Using the variation of ST deferred revenue ignores which portion of this revenue is generated from new versus which is transferred from long-term deferred and which is from renewals. This will yield incorrect conclusions in terms of revenue growth.
- Exclude the impact of professional services billings. If there are Professional Services revenues that are deferred, these will generally only be in short-term deferred revenue. These Professional Services deferred revenue will, therefore, represent a greater portion of short-term deferred revenue than total deferred revenue and will, therefore, have a disproportionate impact on the predictability of subscription revenue.
- Revenue guidance: Almost all software subscription companies regularly provide revenue guidance for future quarters and the fiscal year. Recognized revenue removes pre-billed duration impact, normalizes for professional services and adjusts for timing and weight of renewable billings.
In short, calculated billings can only be a good indicator of future performance if all other factors (i.e. pre-billed subscription duration, the weight of professional services in total revenue, the timing of renewals, etc.) remain constant, which is rarely the case in the technology industry.
Lastly, given the intense focus from audit firms, management teams, and board members in defining, reviewing and auditing software revenue, it is hard to understand why investors place so much emphasis on the far less rigorous metric of calculated billings to predict the future.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.