The first 4 quarterly $ numbers represent 605 and the later two represent 606.
But really the only two numbers we might be able to correct is first JAN and OCT. In Jan call management said in its 606 remarks “Atlas represented 32% of revenue during the quarter, up from 10% in the fourth quarter last year and up from 21% last quarter.”
606 for Jan doesn’t make an adjustment worth mentioning.
Oct Revenue adjusted for 606 was $71.782M. 21% of that is $15.1M
So if we look at the Atlas qoq dollar additions for the 606 quarters and take out $5M in Jan & Apr for mLabs we get
The +4M addition for the most recent quarter is actually fairly strong in regards to previous quarterly performances. It would be twice the comparable quarter 605 qoq growth for example. So the question is why was Q4 stronger and can that be expected to be a stronger than normal quarter in the future (seasonality).
Management comments from Q4
“Our revenue performance was strong across the board and also benefits from stronger than expected Atlas consumption including a number of customers who consumed well in excess of their contractual run rates.”
Atlas customers enter into contractual run rates that mean they contract to spend a minimum amount over the contract period to host applications on the service. Atlas is priced on consumption. From 10-Q.
“We derive our revenue from two sources: (1) sales of subscriptions, including term license and post-contract customer support (“PCS”), and consumption-based database-as-a-service offerings”. The latter being Atlas.
“MongoDB Atlas, our hosted DBaaS offering which we run and manage in the public cloud. MongoDB Atlas provides customers with an elastic, managed offering that includes automated provisioning”
Atlas is consumption based and is (lower case e) elastic with automated provisioning. So the customers are contracted to spend(consume) a minimum amount and as needed the service expands elastically to provision (consume) more.
So what made Mongo’s Atlas customers automatically expand the consumption of Atlas provisions (more nodes and compute power and data)?
I have a hypothesis. This particular quarter encompassed November 1 to January 31. How many commerce related applications are running on Atlas? We’re not just talking e-commerce but all logistics related to a busy holiday quarter. Shipping, financial services, inventory, payroll, customer service, etc. Probably a long list of business applications I imagine.
So yes seasonality seems to be a player. Management said as much and because Atlas was only at $5M/quarter rate a year ago and was $27M/quarter a year later they don’t have a good grasp on exactly how much to expect and are probably therefore not using it to formulate formal guidance yet.
From that Q4 call.
“Clearly, there’s a little bit of seasonality in our business, and so the fourth quarter tends to be a big quarter in general…
So Atlas benefited from a number of dynamics, and we’re very, very pleased with the quarter. Obviously we have the addition of mLab for the first time in the quarter. But even if you strip that out, it was a very strong and accelerating quarter from an overall Atlas performance.
It’s strong edition of new customers. It’s consumption with existing customers. The other thing that we did try and call out in the prepared remarks though was that we did benefit from a number of customers in Q4 who were consuming Atlas at rates in excess of their contracted amounts. And so we always keep an eye to that in terms of like what is the relative run rate and how does that compare relative to the contracted rate.
And so we don’t have enough history with Atlas yet to know as to how much of that is sort of pulling forward or actually establishing a new run rate level. And so we want to keep an eye on that and want to make sure that sort of expectations don’t go out of control as you all extrapolate things, because I do think that that’s a dynamic that’s at play. But it was really very, very strong across the board.”
Clearly there was seasonality in Q4. Adjust for that and I bet Atlas is crushing it. Because if Some customers temporarily increased spend for Q4, Atlas still achieved all that seasonality increase plus another +$4M in this most recent quarter. If my theory is correct then this coming year will see a similr Atlas seasonal expansion due to busy holiday compute requirements and it would be larger than last year due to Atlas being larger.
Unless someone has a better theory.
Just my thoughts.