Calculated Billings

Q2 billings was also negatively impacted by our strong upsell activity. That may sound counterintuitive so let me explain. When large enterprise customers expand their purchases with us, they often want to align new contract start dates with existing contracts, and thus, we only bill them for the stub period before the new large contract starts. This delay of the new contract billing creates a headwind on that quarter’s billings duration, but the new contract ultimately becomes a tailwind to future quarters because of the increased contract size.

That was the CFO of OKTA explaining why their Calulated billing grew only 28% y/y.

So, Slack also had some really big upsells in the first half, increasing customers purchasing more than $1million Annually 78% right?

I am not an accountant. Given most reports on SA state reason for down stock price is fall in Calculated Billings. CC stated $11million In concession or 10% of Calc Billings as the biggest part of that. But…
Does anyone think the same reasoning OKTA gave applies to Slack?



1 Like

Okta’s RPO last quarter was $1.43 billion on $200 million in revenue. (…)

Slack’s RPO last quarter was $388 million on $215 million in revenue. (…)

In other words, Okta has a lot more guaranteed revenue in the future – almost 4x more – on roughly the same amount of quarterly revenue.

Below is how I read Slack’s Calculated Billings, and my reasoning for selling out of my 5% position. In short, there is less certainty for their future revenue than with other SAAS companies. Happy to receive any feedback or corrections on anything below.

First off, a few definitions, as a refresher, for my edification

Remaining Performance Obligations (RPO) represent agreements of one year or longer. Any agreement shorter than one year is not recognized as RPO. For Slack, they disclose in the 10Q that as of the end of the quarter, this was $388.5 million, of which 57% is expected to be recognized as revenue in the next twelve months following 31 July 2020. Remember, under ASC 606, they must recognize 40% upfront and the rest ratably (evenly disbursed over remaining quarters).

Deferred revenue refers to advance payments a company receives for products or services that are to be delivered or performed in the future. In Slack’s case, when they sign an RPO, the revenue that is recognized ratably becomes Deferred Revenue. In other words, most of their Deferred Revenue comes through agreements of one-year or longer.

“Backlog” is revenue that is part of RPO, but has not yet been collected as Deferred Revenue. In other words, RPO = Deferred Revenue + Backlog. Or Backlog = RPO – Deferred Revenue. Slack at the end of the current quarter had RPO of 388.5 million and Deferred Revenue of 383.4 million, so their “Backlog” was $5.1 million.

Calculated Billings is the net change in Deferred Revenue PLUS revenue in the current quarter. For Slack this quarter, it looked like this:

215,864,000 Revenue
383,407,000 Add: Total deferred revenue, end of period
381,073,000 Less: Total deferred revenue, beginning of period

218,198,000 Calculated Billings

This was the number the market didn’t like – “it only grew 25% yoy”. But remember, as far as I know, most of their deferred revenue comes from signing “long-term” contracts of one year or more.

Some further math and my reasoning …

RPO grew 80% yoy. Deferred revenue grew 33% yoy. What does this mean?

If RPO grew 80%, shouldn’t Deferred Revenue also grow by something close to that amount? If it didn’t, shouldn’t “Backlog” then? Let’s look at “Backlog”.

“Backlog” this quarter was $5.1 million. What was it 1 year ago? It looked like this 4 quarters ago:

$215 million - RPO
$285 million - Deferred Revenue

Therefore Backlog was 215 - 285 = -70 million. So Backlog went from minus 70 to 5 million yoy

That is tricky to calculate as a rate of change because it went from negative to positive (if anyone has a way to calculate rate of change from a negative to a positive number, please share). Either way, the difference is 75 million dollars, and has moved in the right direction. This quarter, RPO grew 80% yoy and “backlog” went up a bunch too (we just can’t calculate the rate of change)

(Digression: I also don’t understand how Slack had negative “Backlog”. Maybe they were receiving payments for contracts that were less than one year, while signing fewer contracts for longer than one-year as RPO.)

What does all this mean? To me, it just means they are signing short-term contracts. It’s uncertain! It means we have uncertainty of what revenue is going to look like in the future. They gave guidance, and it wasn’t great.

Also because Slack has to recognize 40% of newly signed RPOs in the current quarter, it means they recognized 40% of that RPO growth of 80% this quarter. Now they will recognize the remaining 60% over future quarters for the life of the contracts. Unless their RPO growth continues to grow at some high number, it’s unsustainable.

We know that Covid19 is hurting their business. We know that Calculated Billings is a lumpy number. Maybe this most recent quarter was the “bottom” of their troubles and things will turnaround. Maybe they will announce stellar numbers next quarter. And maybe their RPO will continue to grow at a high rate. For me though, that narrative doesn’t fit my investment strategy. I don’t do “turnarounds” or “soon it will get better”. That’s just me though. Maybe I’ll be wrong. It doesn’t have anything to do with the narrative that “Slack is a good or useful product”. I do believe it is a good and useful product.

Ultimately, I think you have to believe that revenue growth and operating leverage will continue, without seeing recurring revenue booked ahead of time. One of a SAAS company’s greatest strength is ARR. So if a SAAS company doesn’t have it, then it makes sense that it would be relatively “under-valued” compared to other companies that do have it (like Okta).


So, I think there’s more devils in the details of ASC 606 than many might think.

For example, contracts signed which provide discounted terms for “seats” to be added at a future date are rolled into the RPO. You can think of them as being similar to options contracts or non-cash employee compensation, but there’s a value that may or may not be exercised during the contract by the customer.

So that means, to me, that OKTA could have quite a bit of fluff in the RPO if they negotiate additional seats/licenses aggressively while signing up new customers. (Not saying anything about OKTA except what is ‘possible’ under ASC 606, and that a large RPO provides more ways in which different factors are being rolled into the number. It’s a mistake for us to believe that a large RPO will all drop to the income statement, imo.)

Slack, otoh, doesn’t really have much RPO so I think we all agree they’re mostly operating on month-to-month payments or 1 year upfront subscriptions. (Since RPO > 1 year an annual subscription goes to “current liabilities: deferred revenues” on the balance sheet.) ((A 2 year upfront subscription would put half the payment to RPO, though, and similarly the “same” half the payment would go to “long-term liabilities: deferred revenues”))

Slack does not appear to have great stickiness but this may well be a function of their sales and marketing department. If 3 years from now we find that most customers have been renewing all along we’ll know they are sticky, they just don’t (or didn’t) require longer contracts to get in the door. Also, since RPO includes anticipated or reasonably expected volume usage charges, Slack doesn’t appear to have many of those, either.


That was supposed to go out email. I’m not sure if that was taken from SA Pro. I think I even pasted some from, Bert’s paid service. I’ll ask to have it taken down.