More Mongo

Investment Analysis Clubs Saul’s Investing Discussions
1

I have a real tough time reading reports during the week. Little thing called a demanding job gets in the way! Anyway, I did read some of our own analyses on Mongo during the week ---- thanks as always for that, team! I did read the earnings report in detail tonight. No wonder the stock skyrocketed this week! Some notes that I particularly liked. This is mostly cut and paste from Dev ---- w my own limited commentary in italics:

  1. Atlas revenue grew more than 400% year-over-year and now represents 34% of revenue. And we ended the quarter with over 13,400 customers, up approximately 130% compared to a year ago.
  2. I have to remind people that 98% of the market today is built on relational databases that technology that’s over 40 years old. Look I know that CEO’s often get excited and exaggerate about TAMs, but folks, this number is the real deal. There’s so much legacy crap out there it’s unbelievable.
  3. The total number of MongoDB downloads from our website alone is now more than 60 million with more than 20 million occurred in the last 12 months, up from more than 12 million in fiscal 2018. It should be noted that there are between 20 million to 25 million developers in the world today. . Read that one again folks. What does that mean? Like TWLO, they are really penetrating hearts and minds of developers. And a whole crap ton of them. I know that actual reach to developers can be debated from one person who knows Mongo to another person that knows a competitor…yet THOSE numbers from Mongo are GINORMOUS and real. Follow the customers…
  4. It is very important for investors to understand that since we own all our IP, unlike other open source companies, we have full control over how we license our software.
  5. IBM partnership. Gotta wonder. Is there an eventual deal here? IBM did buy another open source company in Red Hat. They’ve also been busy buying other modern tech companies as well. Who knows. But not out of the question
  6. Our net AR expansion rate in the fourth quarter remained above 120% for the 16th consecutive quarter.
  7. We ended the quarter with 557 customers with at least $100,000 in ARR and annualized MRR which is up from 490 in the third quarter and 354 in a year ago period.
  8. We also saw significant increase in customers spending more than $1 million annually, which increased from 22 last year to 39 this year. That’s another HOLY CRAP. Per Q&A, sounds like some of those are Atlas deals
  9. Gross profit in the fourth quarter was $61 million, representing a gross margin of 71% compared to 76% in a year ago period… As expected, gross margin in the quarter was negatively impacted by mLab, which has a lower gross margin than Atlas. This impact will dissipate by the end of fiscal 2020 as we migrate mLab customers to Atlas.
  10. Document DB Announcement from Amazon. We see no impact, Brad. In fact, I think it’s frankly raised MongoDB’s awareness.
  11. So, Atlas benefited from a number of dynamics and we’re very, very pleased with the quarter. Obviously, we had the addition of mLab for the first time in the quarter. But even if you skip that out, it was a very strong and accelerating quarter from an overall Atlas performance.
44 Likes
2

JKB,

Back out the mLab numbers and the increase in revenues from 606 accounting and instead use 605 to get apples to apples organic numbers. The results are still great but nothing like the euphoria we are seeing. I did that on the NPI board the last two days and did an apple to apple comparison with Elastic today.

If Steppenwulf is correct it won’t matter in the end as the acceleration that was an illusion this quarter due to accounting methodology change and commingling in mLab will become reality in the next few quarters anyways as Mongo crosses the chasm (which is exactly what Steppenwulf was describing).

It does cause me to step down my trust/respect of management however as they went out of their way at the call to commingle this all without going apples to apples.

Compare this to Alteryx. They gave us the 605 and 606 number (~$250 revs 606 vs ~$200 605! - nearly 25% upside difference!!!). With Mongo the accounting change was more like 14-17% exaggeration from the accounting methodology change and another 7% or so for the mLab commingling. Thus at least 21% upside exaggeration if one were to compare organic growth apples to apples. And not even Mongo considers the mLab customers worth much at all economically. Much more of a drag for them. So don’t tell me we should also include mLab as a positive customer and revenue wise.

So apples to apples the number was really materially lower. Fortunately for Mongo, as I indicated, I expect there to be real euphoria quarters in the future showing accelerating growth for real apples to apples as Mongo crosses the chasm. But given these accounting methods, and given that mLabs will now be a comparable next year (and mLab customers ain’t gonna grow) the numbers may actually understate the real business next year. TBD.

Tinker

22 Likes
3

Back out the mLab numbers and the increase in revenues from 606 accounting and instead use 605 to get apples to apples organic numbers.

If one should explain this accounting change related 605/606-issue to their grandmother in layman’s terms, what would one say? I know that I have missed something here, but not sure what it is.

1 Like
4

If one should explain this accounting change related 605/606-issue to their grandmother in layman’s terms, what would one say?

If I were Mark Twain I might start by saying: “There are white lies, damn lies, and accounting.”

If a business is a good business, this is just noise, good to be aware of it so as not to make foolish trades.

Denny Schlesinger

8 Likes
5

If one should explain this accounting change related 605/606-issue to their grandmother in layman’s terms, what would one say? I know that I have missed something here, but not sure what it is.

Without providing links and to make it very brief…

Many companies sign 1 to 3 year contracts. Customers typically pay upfront. Under 605, the company recognized that total revenue on a ratable basis. In the case of a 3 year (36 month) contract, the company recognized the TCV (total contract value) divided by 36 on a monthly basis (or 12 on a quarterly basis).

Under 606, a larger percentage is recognized upfront and a much smaller is ratable over the contract period.

That is it in a gist and if it doesn’t make sense feel free to email me.

A.J.

8 Likes
6

To add to what A.J. said, the rational is that much of a contracts costs are sunk in the beginning, therefore the logic of recognizing more than a ratable ( in his example 1/36 of the contract) part of the income at the beginning. Those costs would be sales costs including the commission or bonus paid to the sales team for landing the contract. Personally, I think the new rules just make it more confusing to compare.

1 Like
7

Tinker -

I disagree with your take on the numbers less mLab and 605 as not being in the blow out / acceleration category. In fact I think we should be euphoric. I will outline my case below :-).

But first… I have seen it hinted at in several posts across the board that somehow the affect of 605 vs 606 may be manipulated, etc. by a company or that a bigger 605 to 606 increase is desirable vs a smaller increase - both of these are not true. These are firm accounting standards that are signed off on in their earnings by their auditors and any differences in the affect of 605 vs 606 in a particular company is simply because of different business models. In an over simplified view, the lower the affect of 605 to 606 transition on a companies revenue actually means the company is MORE SaaS / cloud based than another all things being equal as SaaS is still recognized over time under 606.

Now to Mongo’s Q4 - all data under 605 and apples to apples. First overall revenue:

As reported: $83.1M Q4’19 vs $45.0M Q4’18 → 85% Y/Y

w/o mLab: $77.1M Q4’19 vs $45.0M Q4’18 → 71% Y/Y

I assumed $6M mLab as they guided $5M and said they had moderate over-performance

71% Y/Y growth is the highest growth quarter MDB has EVER had in available 605 history! The next closest was 65% way back in Q1’17 and the TTM revenue growth in the most recent four quarters was 55% Y/Y. This is significantly faster growth than ZS and AYX reported in their respective Q4s.

Now Atlas:

As reported: $28.3M Q4’19 vs $5.0M Q4’18 → 470% Y/Y

Note they give % of overall revenue not an exact number but these are accurate within several hundred thousand dollars

w/o mLab: $22.3M Q4’19 vs $5.0M Q4’18 → 350% Y/Y

The 350% w/o mLab is an acceleration vs ~330% Y/Y in Q3’19 and is by far their biggest Q/Q $ increase of ~$8M (again, excluding mLab).

The non-Atlas piece of the business grew ~37% Y/Y.

I also got the exact opposite impression as you regarding this statement: “It does cause me to step down my trust/respect of management however as they went out of their way at the call to commingle this all without going apples to apples.” → they provided ALL the necessary data! I can give you growth and revenue cuts across 605 and 606 for everything with and without mLab and Atlas vs non-Atlas for the past five quarters based on this earnings release alone. What else could they have provided?

I am also confused by this statement: “Compare this to Alteryx. They gave us the 605 and 606 number (~$250 revs 606 vs ~$200 605! - nearly 25% upside difference!!!). With Mongo the accounting change was more like 14-17% exaggeration from the accounting methodology change and another 7% or so for the mLab commingling.” - MDB gave historicals in both views; all the growth numbers I have shown above are apples to apples. Also as stated above the “bump” from 606 being bigger for AYX may seem like a plus but I consider it neutral or even a slight (very slight!) negative as it means they have less cloud and more on prem.

I have high confidence in AYX, ZS, and MDB and I see no evidence of “fluff” in MDB’s numbers. A company must report their y/y including their acquisitions (TWLO did it last quarter with SEND, SQ is doing it with Weebly) so it is not disingenuous to include it in the topline number and then give everyone the detail to with and without in the release details.

Regards,

Erik

48 Likes
8

Erik,

I agree, they provided all data.

I think Tinker is taking a view that management commentary should have focused on ex mLab rather than all encompassing numbers. I don’t dis-agree, I would have preferred that way, but clearly they have provided all data and more importantly, I think ex-mLab, ASC605 looks fantastic, blow out the top results.

1 Like
9

The revenue growth for Mongo was really good.

The Gross margin drop from 76% to 71% was concerning to me.

I see the gross margin as a a proxy for the pricing power of the business.

During the conference call, they said some of the drop was because of Mlab. Looking at the numbers, because mlab is such a small part of the overall business, it has to be more.

The only explanation is Atlas is lower margin than their other business. If so, as more of their business becomes Atlas, the gross margins will continue to drop.

I think this is a big deal, they will get less profit for each dollar of sales in the future.

Why does no one else think this is a big deal? Just care about revenue growth?

Thanks,
Jim

4 Likes
10

Jimbo,

The reason for the margin going from 76-71 is because of two simultaneous things that have occurred.

  1. Atlas has lower gross margins than Community Enterprise subscriptions. SaaS is always less gross margin than subscription software because it is hosted. Particularly for a database product where a lot of nodes need to exist to host peoples data and infrastructure to process. Atlas growing so fast and into an ever larger portion of revenue will work to lower margins. But the margins are still high and probably not much less than that 70 range.

  2. The bigger issue for this quarter was mLabs acquisition. This business has even lower gross margins than Atlas, because of the nature of what mLabs was. So even though it’s smaller amount of money the effect is larger. Plus an acquisition like this will initially bring a lot of cost overlaps and duel hostings.

The reason why not many are all that worried can be summed up by this statement from management in the call.

As expected, gross margin in the quarter was negatively impacted by mLab, which has a lower gross margin than Atlas. This impact will dissipate by the end of fiscal 2020 as we migrate mLab customers to Atlas. Overall, we’re very pleased with the continued gross margin improvement of organic Atlas during the fourth quarter. However, we continue to expect that we will see some modest reduction in overall gross margin as Atlas continues to be a bigger portion of our revenue.

Both the pressures are not permanent.

  1. As Atlas continues to scale the gross margins tick up. Not all of the cost to host is duplicated for each new dollar. Those costs of revenue get spread out over more revenue and become a less percentage wise of the whole. Thus over time the gross margin for Atlas rises. I’m not sure what the long term goals are for Atlas gross margins are. A wag would be high 60s low 70s. So even if Atlas hits 50,70 or higher percent of business that’s the floor.

  2. The mLabs thing will work itself out. The cost duplications from any acquisition get negated over time. Further most mLab customers will be migrated over to the higher margin Atlas and see the previous paragraph for that. This should largely be gone as a factor by the end of the year if management is to believed.

Personally, I don’t see much of a problem here with a product like Atlas with such incredible growth and increasing margins and the dynamics with how that effects Mongo’s overall gross margins. Someone posted a link explaining SaaS business metrics. This is a good example of that playing out for Mongo.

Subscription revenue made up over 94% of revenue, with the balance made up of really low margin services. If services was a higher percentage that might be concerning. But as such, it’s not.

Darth

20 Likes
11

The only explanation is Atlas is lower margin than their other business. If so, as more of their business becomes Atlas, the gross margins will continue to drop.

I think this is a big deal, they will get less profit for each dollar of sales in the future.

Why does no one else think this is a big deal? Just care about revenue growth?

Hi Jim,
Well, let’s see: Atlas growing at 300%. Gross Margins 5% lower. Well, you do the math.
Saul

9 Likes
12

Darth-

Thank you for your thoughts.

This is the part I was missing:

  1. As Atlas continues to scale the gross margins tick up. Not all of the cost to host is duplicated for each new dollar. Those costs of revenue get spread out over more revenue and become a less percentage wise of the whole. Thus over time the gross margin for Atlas rises. I’m not sure what the long term goals are for Atlas gross margins are. A wag would be high 60s low 70s. So even if Atlas hits 50,70 or higher percent of business that’s the floor.

It will be interesting to see how big a percent of their business will be Atlas in the future.

Jim

1 Like
13

It will be interesting to see how big a percent of their business will be Atlas in the future.

Also, it will be interesting to see if Dev/his team eventually take ORCLs playbook from the 90’s and expand from database to applications dev. Particularly ERP ---- as cloud ERP still feels overpenetrated by the dinosaurs.

CRM has done really well w applications dev but Benioff/team have purposefully stayed away from building ERP. Kenandy is an option as is Netsuite (ORCL). But I do wonder if Dev/team eventually follow the same playbook from ORCL in the 90’s.

Purely speculation.

14

The Gross margin drop from 76% to 71% was concerning to me.

I see the gross margin as a a proxy for the pricing power of the business.

During the conference call, they said some of the drop was because of Mlab. Looking at the numbers, because mlab is such a small part of the overall business, it has to be more.

The only explanation is Atlas is lower margin than their other business. If so, as more of their business becomes Atlas, the gross margins will continue to drop.

Jim -

I agree gross margins are something to watch and it’s something the company is aware of as well. The CFO made this statement in his prepared remarks for Q2:

Gross profit in the second quarter was $41.8 million, representing a gross margin of 73%, which was slightly better than the year ago period. We are pleased with the gross margin performance in the quarter, particularly in light of Atlas’s growth. Atlas includes the underlying infrastructure and our success in reducing the infrastructure costs related to Atlas have resulted in improved Atlas gross margins. We have now begun to introduce the premium Enterprise Advanced features into Atlas, starting with advanced security and we expect this will be an incremental benefit to revenue and gross margins going forward. We still anticipate Atlas will be a modest headwind to gross margins overall, but we are tracking well relative to our original expectations.

In the Q&A that Q, there was this exchange:

Brent Bracelin

And then as a follow-up for Michael. Maybe if I look at the gross margin profile here, I guess it’s a little surprising that you’re able to kind of hold the gross margins even with the strength that you’re seeing in Atlas. So can you just walk us through with that material mix shift to Atlas, how you’re able to kind of sustain the gross margin profile here.

Michael Gordon

Yes, sure. So as I mentioned earlier, we’ve made very good progress on the Atlas gross margins, relative to our expectations and we’re tracking well ahead of plan in terms of gross margins. Atlas’s still lower gross margin than the balance of the business. And so we still have plenty of levers to flip or turn and dials to adjust. On the business, we mentioned that the first thing that we had done which we’d said we’d do by the end of the year is we’ve introduced some of the premium features and Enterprise Advanced, meaning principally the advanced security features into availability in Atlas.

We’ve not yet seen the benefit of that, because we only just introduced that. But over time, that should be very accretive both from a revenue perspective as well as from a gross margin perspective. And there are some internal cost optimizations that are also available to us. So, I think the short answer is, we’ve had very strong execution relative to our kind of margin game plan for Atlas, but we’ve got more work that we want to do on that front, but we’re very pleased with the progress to-date.

So Atlas is indeed lower margin, but it’s something the company seems to think should improve as it grows and premium features are added. I have their gross margin numbers as follows, but the change to 606 might very well screw them up royally since I only have 606 non-GAAP rates for 4Q18, 4Q19, FY18 and FY19:


non-GAAP Gross Margin					
	Q1	Q2	Q3	Q4	YR
2017	67.8%	68.8%	72.6%	72.1%	71.6%
2018	71.6%	72.0%	74.1%	76.0%	74.9%
2019	73.0%	73.0%	74.0%	71.4%	73.7%

Without true 606 comps (Help? Anyone? Bueller?), I’m not exactly sure how to weight it. That being said, their other numbers and the market’s reaction to them don’t have me overly concerned just yet. As you said though, this is something to continue tracking as Atlas becomes an ever-bigger part of their business model.

1 Like
15

…or just read what Darth and Saul said about the Atlas margins.

I guess that’s what happens when you fall 60 posts behind.

16

More thoughts on the dynamics involving Atlas margins and the cost of hosting.

Atlas is hosted on AWS, Azure, or Google Cloud. Mongo has to pay those companies to host Mongo’s customers data and to perform compute instances on that data. Using cloud storage and compute resources and other resources provided by those providers. This expense is variable depending on the the amount of resources needed to provide the Atlas service. More customers, more data to store, more data to process, more cost. Of course Atlas passes on the cost of those instances with Mongo software as the value add on. As the size of the customer’s need goes up and they spend more (cost of instance + Mongo value add) the gap between the total of what they are spending and the cost of the instance to Mongo widens. And Mongo sells a number of Atlas add ons and services that produce more revenue per use. Bigger deployments (higher average customer value), which Mongo is achieving, increases margins.

Making up numbers here. A customer who spends a $1 on an Atlas instance might result in a charge from AWS of $0.26 to Mongo to perform the instance. Another customer who spends $10 on Atlas might generate a charge of $2.50 to Mongo. Not from an actual example, but just trying to highlight how increased spend on Atlas creates more cost to Mongo, but scale can have a variable effect on gross margins.

Other areas where growth leads to better margins are the pricing structures of cloud providers. Mongo gets better rates per instance as they consume more. In addition, increased spend to the cloud providers provides them better leverage for negotiating better rates. All of these dynamics work to improve Atlas margins over time and scale.

Headcount is another cost of revenue. You will need to hire employees to manage the service and provide support. This is another area where as the business scales, sure you have to hire more employees to perform these functions, but it’s not a one for one thing. This effect is lessened as the SaaS business expands with the biggest impact in the beginning. It will always go up as you grow but not always at an equivalent rate. The gross margin impact decreases as scale and time increase.

Other cost are not quite directly correlated to scale. This is the cost of the software. Typically ammoratized over time. You spend to create the software. You spend to update it. But it doesn’t go up as revenue increases. And the original cost to develop the software fades and the cost to maintain and update software is not a constant. Some periods it may be higher if there’s a big update and other times less. But over time it’s probably fairly flat. So this type of expense again has an improving margin impact with time and scale.

Other costs for hosting include things like credit card transaction fees and other things that are more common to any business model.

You can see how big clients with lots of spend are so important to SaaS companies and worth spending a lot on S & M to acquire. And hopefully that helps make the case why SaaS will typically have lower margins than subscription on premises software particularly in the early stages and how those margins improve with scale and time. But there is a ceiling and it will be lower than what a non-SaaS company like AYX can achieve.

Darth

14 Likes
17

I’d tell granny the accounting change makes the company count revenue sooner than the old way. They still make the same amount of money either way, the change just makes it current revenue sooner. The old way called in “deferred revenue” until it came time to enter it as current.

1 Like
18

Personally, I can’t see any new ERP product becoming profitable. There’s just way too much established competition. Why would you buy a new offering when there are so many established, fully functional applications out there?

Furthermore, ERP as a use case is very well suited to a relational implementation. Lost of discrete well defined data, it is practically a description of an entity-attribute data model. CRM is entirely different, dealing with image, text document, email, presentation, message, voice, in fact just about any and all data types are part of the CRM application data. Perfect use case for no-SQL database.

1 Like