Datadog earnings and CC - My summary

My summary of the Datadog press release of earnings and their conference call… Please note that I edited, shortened and paraphrased extensively so that anything that doesn’t make sense is probably my fault.
Saul

Revenue grew 88% to $96 million. (Note that it’s ALL subscription revenue)
Customers with an annual revenue rate of $100,000 or more, grew 93% to 727 from 377 a year ago
Announced over 15 new products and functionalities at annual user conference
Raised $709 million through IPO

Datadog, the monitoring and analytics platform for developers, IT operations teams and business users in the cloud age:

We are very pleased with our quarter, highlighted by 88% yoy revenue growth, and continued traction with larger customers. Datadog has established itself as the leading monitoring and analytics platform and we have continued to extend our capabilities during the quarter. We announced over fifteen new products and functionalities at our annual user conference in July, including Network Performance Monitoring and Real User Monitoring.

Financial Highlights:
Revenue was up 88%.
GAAP Gross margins were 76%
Adj operating income was $0.6 million;
Adj operating margin was 0.7%.
Adj EPS was 0 cents.
Operating cash flow was $3.8 million,
Free cash flow was $(3.7) million (due to realestate capex expenses).
Cash, was $771 million.

• In July, we hosted our annual user conference, Dash, which was attended by more than 1,200 people in NYC. We introduced over fifteen new products and functionalities, including Network Performance Monitoring, Real User Monitoring, Serverless Functions, Logs Rehydration, and more. These new products and features extend us further on our path to providing full-stack visibility to our customers, all in a tightly integrated platform.

• In September, we completed our IPO of 27.6 million shares at $27 per share, for total net proceeds of $709 million.

• In October, we announced we are currently “In Process” on the FedRAMP Marketplace, initiating the certification process. Achieving FedRAMP authorization will expand our addressable market, by allowing U.S. federal government departments and agencies to adopt and use ourcloud platform.

We were recognized by Forrester Research as a Leader in The Forrester Wave: Intelligent Application and Service Monitoring, Q2 2019. For this report, Forrester evaluated 13 vendors across three main categories: current offering, strategy, and market presence. Datadog scored highest among all vendors for strategy.

Conference Call

The CEO: My co-founder, Alexis and I started Datadog about nine years ago with a mission to break down silos between developers and IT operations teams. Today, we are THE monitoring and analytics platform for dev, ops and business users. We provide clarity and actionable insights into software applications and IT infrastructure all in real time. We exist so that our customers can understand everything that happens in their technology stack. While we are very proud of what we have built, we’re even more excited about the future and the tremendous opportunity ahead of us.

A quick review of our results. Revenue was $96 million, up 88%. Adj operating income was $638,000, for a margin of 0.7%. We ended the quarter with 727 customers with annual run rate revenue, or ARR, of $100,000 or more, up 93% from a year ago.

As in past quarters, our dollar-based net retention rate was over 130%, as customers increased our usage and adopted our new products. And we continue to be capital efficient as our cash provided by operations was $3.8 million and year-to-date $6.8 million.

As you all know, a massive IT platforming is underway. Companies are moving from static on-premise architecture to public and private cloud as well as other ephemeral technologies like containers, microservices and serverless computing. These newer technologies allow for increased agility and innovation, but also compound complexity. Developers and IT operations teams which used to be separate must come together in order to manage IT chaos and better collaborate around a shared view of the IT stack.

These challenges are affecting companies across all industries, geographies and sizes. We believe we are at the very early stages of an existential market opportunity, which we estimate to be approximately $35 billion.

Datadog was founded in 2010 as a real time data integration platform that turns the chaos from disparate sources into digestible and actionable insights. Our vision was a single platform that would provide dev and ops users with a common view across sources, teams, and technologies.

In 2012, we launched our initial use case infrastructure monitoring. Since initial launch, we have continued to innovate across more environments including containers and serverless as well as on-premise hybrid product in multi-cloud environments.

We saw a need for full-stack observability and we launched Datadog Application Performance Management or APM in 2017. We quickly added Datadog Log Management in 2018 thus completing what we like to call the three pillars of observability. Earlier this year we also launched Datadog Synthetics to extend it to user experience monitoring by letting our customers simulate user journeys on their web applications in API endpoints.

All our functionalities are offered within the same tightly integrated platform. Our customers can frictionlessly add-up new products or from this end-user interface, powered by a common data model. We win in the market for several reasons. One we are a truly integrated platform allowing us to solve our customers end-to-end problems and innovate rapidly. Two, we were built for the modern dynamic stack offering end-to-end visibility. Three, we are simple but not simplistic, easy to install with no professional services. And Four, we are designed for use in collaboration across development operations and business teams.

From a business perspective, we have an efficient operating model which has enabled us to have huge growth with very modest cash burn. Despite significant and ongoing investment in R&D and S&M we have only burned approximately $30 million in cash since we began. We had a very strong cash pay back. With that in mind, let’s review our third quarter performance.

Overall, we are very pleased with our results. Strength was broad-based driven by both robust new logo additions as well as continued growth of existing customers. Our platform strategy is clearly resonating including strong initial uptake of our Synthetics product. From the R&D perspective, we continue to invest in our product suite. And we announced over 15 new products, features and functionalities this July at Dash, our annual user conference. One of the products we announced was Network Performance Monitoring to allow customers to visualize the flow of network traffic in all environments. It is an extremely lightweight solution that is compatible with all major cloud providers and on-premise servers, giving customers the flexibility to monitor network traffic without sacrificing performance.
Another new product was Real User Monitoring, which complements synthetics for user experience monitoring. It allows customers to analyze the performance of applications directly experienced by the end users. Additionally, we announced enhancements to our machine learning and AI capabilities. Watchdog, Metric Coalition, Trace Outliers, etc.

Last but not least, is that we are in process for FedRAMP certification. We’re very excited by the potential to extend our market to Federal govt departments and agencies.

As a quick note, two products were announced at Dash in beta, and we are not yet charging our customers for them. We began charging for Synthetics this quarter.

We saw strong adoption of our newer products from both new and existing customers. About 50% of our customers were using two or more products, up from 40% sequentially and 15% yoy. And our newer products are no more than about 2.5 years old.

We plan to continue to invest in R&D. Year-to-date Adj R&D expense was 30% of revenue, up from 27% last year.

We had approximately 9,500 customers up from 7,100 a year ago. We ended with 727 customers with an ARR of $100,000 or more, up 93% from year ago, an increase of more than 130 in the quarter alone. Given that more than 70% of our ARR is generated from customers over $100,000, we expect this cohort of customers to be a large driver of our future growth.

We continue to invest in our S&M. This includes growing sales reps by 70% yoy. We get high returns on our S&M investments benefiting from our very efficient business model and driven by our land-and-expand model. As evidence of our efficiency, our cash payback continues to be about one year so we intend to continue investing meaningfully in our S&M efforts globally.

The CFO: We generate revenue from the sale of subscriptions to our SaaS platform. Customers have the option to purchase multiple products including infrastructure monitoring, APM and logs as well as additional SKUs such as containers, custom metric packages, and anomaly detection. Our revenue is all subscription as professional services are not required to implement our products. Customer contracts typically have either annual or monthly commitments. Additionally, customers are billed for on-demand usage in excess of their committed amount, typically monthly in arrears.

Given the mix of annual and monthly invoicing and the variability in billing, we don’t believe that calculated billings is a very useful metric for investors to evaluate our business. In any one period, billings growth can vary substantially from revenue growth.

Turning to Q3 results, revenue was $96 million, up 88%. The strength was broad based, driven by new and existing customers as well as strong platform adoption and initial uptake of Synthetics. This is particularly notable in the seasonally challenging summer quarter.

Our dollar-based net retention rate was above 130% for the ninth consecutive quarter. A robust retention rate is driven by increased usage of existing products as well as cross-selling of newer products, including the inclusion of Synthetics for the first time. Once our customers are on the Datadog platform, they can frictionlessly expand with us to increase use of the platform. The cross-selling of newer products is a more recent driver of our net retention rate. Lastly, we’re now showing strength internationally, specifically a strong sales quarter in EMEA. This is particularly impressive given many of the teams in the region are still ramping.

Calculated billings was $112 million, up 132% yoy. We don’t think billings should be used to evaluate our business due to the mix of monthly and annual billing terms among our customers. And because billings growth can vary substantially from revenue growth in any one period, nor do we plan to comment on a regular basis going forward.

As a reminder, unless otherwise noted, all metrics are non-GAAP. Gross profit in the quarter was $73 million, representing a gross margin of 76%. This compares to a gross margin of 77% a year ago and 75% last quarter. We are now in the middle of an accelerated innovation cycle of delivering new products, as well as the build out of cloud data centers in newer geographies. We will over time balance investment and optimization to manage our gross margin.

R&D expense was $27 million or 28% of revenue, up slightly from 27% a year ago. We have made extensive investments in our platform. This is shown by the launch of Synthetics earlier in the year and the 15 plus product announcements made it Dash.

We continue to see a meaningful opportunity to further our innovation and expand our platform, and therefore plan to continue to make meaningful investments in R&D going forward.

S&M expense was $37 million or 39% of revenues down from 48% a year ago. The change was more pronounced than usual due to the timing of trade show events in 2019 versus 2018. For nine months, S&M was 41% of revenue, down from 43% a year ago showing some leverage in S&M.

We continue to see strong returns from our S&M investments and plan to continue to invest to expand our go-to-market globally. G&A expense was $8.2 million or 9% of revenues slightly higher than 8% a year ago given some IPO related expenses. Operating income was a positive $638,000 or a 0.7% operating margin compared to an operating loss of $3.2 million or negative 6% in the year ago period.
Net income for the quarter was $695,000 or breakeven per share. We have a highly efficient business model and experience a high return on our investments in S&M and R&D. We see ample opportunities to continue to invest in the large market opportunity ahead of us.

Turning to the balance sheet and cash flow, we had $771 million in cash. This includes $709 million in IPO proceeds. Cash flow from operations was positive $3.8 million for the quarter and positive $6.8 million year-to-date. Free cash flow was a negative $3.7 million for the quarter and a negative $10.1 million year-to-date.

Our capital expenditures consist primarily of real estate build-outs and was slightly elevated due to projects in our New York and Paris offices. The timing of some of the payments for these build-outs shifted to Q4 and we therefore expect another quarter of slightly elevated CapEx. Given we are hosted entirely in the cloud, our hosting costs flow through the P&L, not CapEx.

Second, on profitability. In Q4, we will have the first full quarter of public company costs, particularly D&O insurance (Directors and Officers Liability), as well as some sizeable trade show expenses.

To summarize, we are very pleased with our business performance. We have built the leading monitoring and analytics platform for the cloud age and are generating growth at scale that few companies can match. We are making continued investments for growth in the foreseeable future. We believe we are at the early stages of a multibillion dollar market opportunity and we feel very good about our ability to build a very large and successful company over time.

Question-and-Answer Session
Q - Congrats on a very strong set of results. Did you see strengths coming from competitive displacements? Are you seeing displacing competitors or are you continuing to see a lot of the growth coming just from Greenfield expansions?
A - The business is still mostly Greenfield and it’s mostly net new and it’s mostly cloud environments, which by definition are new for our customers. They all have other solutions, especially, when you talk about large customers and large enterprise customers, they have existing solutions for their legacy environments, but they make new decisions for the cloud environment. We do see also a few competitive displacement, but that’s not the majority of what we do.

Q - You’re seeing some strong momentum with new products. Synthetics has some really good early signs of traction. And you had some additional opportunities in international. So maybe if you could sort of rank some of the opportunities in terms of what are the biggest opportunities that you see over the next one or two years. And what are the some of the more near-term opportunities that you think can drive growth more than near-term.
A - Well, near-term our core infrastructure product is growing very, very fast, and is what we know most of our customers use along with other products, and still grow a lot with that product. In addition to that our APM and log management products are also very recent and they’re in both in hyper-growth. So they drive a lot of our short-term and near-term success.
Newer products like Synthetics are a nice add, but don’t represent much bulk of performance. Right now, it’s still the three pillars, the products we have in the market and we’ve been developing over the past few years, that are getting us the bulk of our growth. We’ve been very happily surprised by the performance of Synthetics right out of the gate. It’s a great sign. It’s a great sign, in particular, when we think of all the new products we have that we haven’t started charging for. But it will be more material in quarters or years to come.

Q - In terms of the customers that are going with multiple products. How much of that success is coming out of the mid market versus large enterprise?
A - We see the same behavior in all parts of the market when it comes to that. In fact, the places where we’ll see customers not using us for everything they do, are those that got into the cloud earlier and have used other products for a while and have everything set up. That’s where we don’t have an initial opportunity to displace everything at once. When customers are newer to the cloud and they’re starting their migration, which is the case of most of the enterprise customers as well as the higher end of the mid market there’s an immediate opportunity to land on two or three or four of our products at once.

Q - The market has seen your pricing that you’re using for logging as very disruptive. And I think we’re starting to see some of the competition drop their pricing. Wondering what you’re seeing in the competitive dynamics with all of those changes.
A - We don’t see any change in the competitive dynamics yet.

Q - Can you just talk a bit about how the ongoing shift to containers has changed the competitive landscape for you at all? Is that an increasingly important factor in determining why you win with customers?
A – Yes, definitely – When you think of what makes the old guard of tooling just not work at all, containers are one big part of that. And we’ve been monitoring containers at scale for many years now. We’ve seen our customers run containers in production since before it was advisable to do it. And that’s been a big part of our in-production monitoring product for many years now. We don’t have numbers to share on the volumes we see, but I can tell you that we see lots of volume.

Q - As you continue to invest in R&D, maybe can you just talk a bit about some of the main projects there? I mean, obviously you have a very strong and comprehensive suite in the market already, but with the incremental R&D investment, are you looking to go broader in terms of new products? Or is it just maybe more focused on deeper in terms of functionality with some of your existing products to give you even further traction in the enterprise?
A - Well, it’s a bit of both. Our infrastructure product is still growing very fast. And there’s a lot we can still do to improve it. So we are heavily investing in our partnership product, on log management and APM products are in hyper-growth and we’re very rapidly innovating there. We’ll have a few more products we’ve announced at Dash that we have to fully bring to market. Most are in beta and we haven’t started charging for them yet. That will begin in the next few quarters. And then there were few larger categories we talked about in the IPO road show that we think we can enter in the future. But that’s I would say mid to long-term plan.

A - At the end of the day, what matters is to make the product as easy to deploy as possible, as easy to get information in and out as possible. And that’s where we invest heavily on the backend of our products.

Q - Can you update us on how we should expect gross margins to scale as you continue to see strength internationally and need to ramp compute capacity overseas?
A - In our long term model, we suggested a slight improvement over time. However, now we’re in investment mode, so we’re balancing investment versus optimization. We’ve been in the range of 75%, 76%. We said that as we get to 75% we’d start to try to balance a little more. But we’re still in investment mode and we’ll still expand geographically.

Q - Congrats from me as well. If I understand you correctly, people come to you when they move over to the cloud, because the old tools, the on-premise tools don’t work anymore… Where are we in that cloud migration? It feels very, very early, and even on infrastructure monitoring, there should be a huge opportunity of initial workloads coming over, but then more and more workloads come in. I’d just like to hear your big picture thinking there.
A - Nobody knows exactly where we are. It’s hard to wrap your arms around the whole thing. But it seems that it’s either a single-digit or a low teens percent of the workloads that have moved from legacy IT to public and private cloud. So, with that in mind, we have basically 95% or 90% of the end market still in front of us, which is why we think even for infrastructure, which was our first product, there’s a tremendous amount of runway ahead of us.
Also, as companies transform and become digital businesses, their applications and software are going to grow. We’re not just moving from what it was before into something that’s more in the cloud but the same size. The end state is going to be a lot larger. It’s hard to quantify, but we are confident that this is a very large opportunity and we are the very, very, very beginning of it.

Q - I’ll offer my congrats as well… When we talked to folks out in the industry, many people think that maybe only 5% of apps are being monitored. I’m curious, I guess why is that? And five years from now, where do you think we could be at as an industry?
A - You are talking about legacy APM (Application Performance Management) that was used in the traditional data center apps. The reason for the low rate of use is that these legacy APMs are very, very, very heavy weight and they’re very expensive. It’s very hard actually to deploy them and get value out of them and it ends up being limited to a small set of extremely high value apps, for which you can be convinced to make an investment and get some ROI out of it.
When you think of the world of the cloud, the world of companies that are becoming increasingly software companies, they’re going to have many, many, many, many more apps. And the solutions we’re providing to them are a lot easier to deploy and it’s actually a lot more affordable for each unit of compute. So, we’re going to end up with a market that is significantly larger and there’s going to be a lot less investment needed to get to see returns. So, that’s the big difference between this world of the 5% of the apps being monitored with APM to the world of the future, where companies will be mostly digital and they will end up monitoring most of the applications.

Q - And then congrats on the early successes of synthetics, I imagine a lot of that greenfield wins, but could you talk a little bit more about the competitive landscape for synthetics?
A – There are a few different aspects to it. First, a number of the APM vendors have a Synthetics capability on top of their APM product. So that’s the first set of competitors. A number of companies offer a pure Synthetics testing. Some are fairly low-end, so some cheap products that you can very easily adopt. And then there’s another part of the industry in a high-end software testing. That’s the overall landscape.
Our initial focus was not to try and compete with any of those. It was to fill a gap we saw in the world customers could understand. So, we didn’t approach it with an intent to compete head-to-head with those companies. We approached it from what we think the future is going to look like in terms of an integrated platform that covers everything including simulating user traffic.

Q - A quick follow-up on Synthetics. You mentioned a strong uptake, could you dive a little deeper into that comment in terms of what incremental revenue you’re seeing. And the second question is: As you continue to expand this product portfolio, which is obviously very robust, is there an easier way that you can go to market with an enterprise license agreement, or some type of suite pricing, that will make it easier for the sales force to go to customers and engage as you roll up more and more features? Are you seeing desire from customers for that or is it still too early? Thank you.
A - On the first part, yes, we’re very impressed by what we feel right out of the gate with Synthetics. I think it’s we can say a couple of percent of the overall growth rate comes from Synthetics, which was more than we had expected. I will say though that this includes a few months of pent-up demand. As we onboarded customers on the beta, we didn’t start charging for the product right away. So, when we started charging this quarter, we already had some buildup usage. Over time, we expect it to grow.
On the pricing, and the way to sell multiple products, we already do that in a way, we let customers commit to a certain amount of spending according to usage on Datadog by any of the products, and they can combine them in any way they see fit, without having to commit to specific level of use for any of the products.
And that’s something our customers love, because it lets them just negotiate a rate for everything ahead of time and not have to worry about what their teams are going to use or not use. And that hasn’t been a problem for the sales force to sell that to customers.

Q - Have you seen any difference from year to year in customer draw down rates on existing contracts?
A - Well, we continue to see the same effect, which is customers are drawing down in both periods more quickly than the pro rata, so they’re consuming the products both from the uses of infrastructure as well as additional product.

Q - Congrats on a great first quarter out of the gate. You guys have pioneered so much in terms of monitoring and looking at this from a multi-siloed approach, there have been some competitive responses from a couple of companies in the APM space and the log space. Do you see any change in sales cycles or decision cycles in response to competition from people following your lead?
A - No, we don’t see any changes. As our approach is obviously successful we do expect that it’s not lost on our competitors. We see that as that validation of our approach, but as of now we haven’t seen any changes in the competitor dynamics.
We had a very strong new logo acquisitions, you can see that in the customer count. And even over the summer months, and we also maintained the high retention rates with increased usage, as well as having an increased adoption of the platform, so all of that is continuing.

Q - Let me add my congratulations. I’d love to hear your thoughts on where customers are in their journey towards DevOps maturity and how you see it differ by geography, or by industry or size of company?
A – If you think of the transition to DevOps related to cloud, the markets in EMEA and APAC are about two or three years behind what we see in the U.S.
The first companies to move to the cloud and to DevOps were the small companies, tech-oriented companies. After that are larger enterprise companies that need to modernize, and I think last are going to be the mid market, more traditional companies that are trying to mobilize as well. So we see all those various parts of the market turn on one after each other basically.

Q - The dollar base net expansion rate and what we should expect as we think about it from the future?
A - We had said IPO that we would comment on that relative to 130. And we’re continuing to do that. We’ve seen no change to what we saw when we went public relative to the drivers of that, increased usage and cross-sell, and in our perspective 130% is already best-in-class and that’s the number we’re going to use moving forward.

Olivier Pomel - In closing, we are incredibly proud of what we built. We believe we’re in the early stage use of a substantial re-platforming opportunity. We are very focused on executing on our growth strategy today and we believe we have the potential to be a much larger and profitable company in the long-term.

84 Likes

Nobody knows exactly where we are. It’s hard to wrap your arms around the whole thing. But it seems that it’s either a single-digit or a low teens percent of the workloads that have moved from legacy IT to public and private cloud. So, with that in mind, we have basically 95% or 90% of the end market still in front of us, which is why we think even for infrastructure, which was our first product, there’s a tremendous amount of runway ahead of us.

Wow. I knew it was early but not THIS early! A lot of SAAS growth ahead!

-AJ

6 Likes

I’m not a techie but I get the impression that the products that DDOG sells are low switching cost/easy to replace with another company. The fact they don’t have any consulting revenue reinforces it. And there are a lot of other competitors out there that could challenge DDOG.

So it’s good that DDOG does not have to spend a lot on S&M for their growth as it may not pay off with customer retention rates.

The viral adaption and easy sell is a bonus because it allows DDOG to grow as quickly as it is. But it could be a double edge sword because a competitor that begins offering a worthy alternative could just as easily start taking market share just as quick.

It seems DDOG has the best product at this time but I think it would help to know what the competitors are doing because this is a crowded space after all.

Can anything be done to achieve this other than the Gartner reports and watching the numbers of competition?

2 Likes

Nobody knows exactly where we are. It’s hard to wrap your arms around the whole thing. But it seems that it’s either a single-digit or a low teens percent of the workloads that have moved from legacy IT to public and private cloud. So, with that in mind, we have basically 95% or 90% of the end market still in front of us, which is why we think even for infrastructure, which was our first product, there’s a tremendous amount of runway ahead of us.

Congratulations, SoonerAJ! You picked out the “S” curve scenario for cloud monitoring technology! At 5 to 10% market penetration it has not even reached the curve in the hockey stick or reached it very, very early. If DataDog truly has no real competition, then it’s truly a fabulous opportunity. As I pointed out some time back, I had to write an error monitoring app several years ago and DataDog is providing just that in spades and wholesale.

Denny Schlesinger

PS: Visibility in the cloud is a strange metaphor but that is exactly what some airplane instruments do like the artificial horizon.

The attitude indicator (AI), formerly known as the gyro horizon or artificial horizon, is a flight instrument that informs the pilot of the aircraft orientation relative to Earth’s horizon, and gives an immediate indication of the smallest orientation change. The miniature aircraft and horizon bar mimic the relationship of the aircraft relative to the actual horizon.[1][2] It is a primary instrument for flight in instrument meteorological conditions.[3][4]

https://en.wikipedia.org/wiki/Attitude_indicator

6 Likes

I’m not a techie but I get the impression that the products that DDOG sells are low switching cost/easy to replace with another company. The fact they don’t have any consulting revenue reinforces it. And there are a lot of other competitors out there that could challenge DDOG.

You are, of course, right, but there are two sides to this scenario. Complexity is the enemy of adoption and the consequences have been felt here with products that have “longer sales cycles.” As you say, simplicity invites competition but why would customers switch? There are reasons for high retention rates. First mover advantage or path dependence is the most important one. If DataDog were a one trick pony your scenario would be reasonable but DataDog is offering a portfolio of monitoring products in a tightly integrated control panel. If the competition uses a different control panel your central monitoring advantage vanishes. This is the reason why DataDog needs an aggressive land and expand sales strategy which the simplicity of the product promotes. This is the reason why one product takes 60% market share and there is a long list of wannabes. One classic example of old was IBM and the Seven dwarves.

Besides IBM, there were lots of players in the build-a-new-computer game for a short while… Digital Equipment, Control Data, General Electric, RCA, Univac, Burroughs, and Honeywell.

https://it.toolbox.com/blogs/shayne-nelson/the-60s-ibm-the-s…

General Electric and RCA were giants in their own right but could not take on IBM in IBM’s area of competence. It got to the point where “You can’t lose your job by hiring IBM.” A screwup with one of the dwarves was career threatening.

My point is that if cloud monitoring were truly a commodity product your argument would be reasonable but, at this stage, cloud monitoring is mission critical and not (yet?) a commodity.

Denny Schlesinger

16 Likes

Saul,

Reading through your summary of DataDog (DDOG), the following statement stood out to me:

As you all know, a massive IT platforming is underway. Companies are moving from static on-premise architecture to public and private cloud as well as other ephemeral technologies like containers, microservices and serverless computing. These newer technologies allow for increased agility and innovation, but also compound complexity. Developers and IT operations teams which used to be separate must come together in order to manage IT chaos and better collaborate around a shared view of the IT stack.

I see a lot of technical jargon in there, yet my technical/software expertise interprets it in an interesting way as an investor. Here is my take on what this means:

Software is rapidly moving from simple desktop applications to the cloud and other new types of software architectures. This shift is rapidly adding complexity and often a bit of chaos. These new software architectures has grown too big to keep track of what is happening using existing tools.

DataDog fills in a vital missing component: Provide a simple view into what is happening with these increasingly complex software systems while maintaining the ability to link the simple (overview) to the complex (details of individual errors). DataDog creates order from chaos.

On the surface, this is nothing new. Many companies have tried to provide insights into the performance of these increasingly complex software applications. Two obvious companies come to mind, both previously discussed on this forum: New Relic (NEWR) and Splunk (SPLK). However, those companies and other competitors to DDOG provide nowhere near the capabilities I see in DataDog.

This indicates to me that DDOG probably has a significant moat. What they have achieved is probably hard to replicate, as attested to the fact that other companies have tried and fallen short

This also indicates to me that DDOG is providing a type of service that is needed in just about every other type of modern software. This in turn indicates DDOG may have far more room to grow than I initially gave them credit for.

All this reminds me of something I read of Warren Buffett. That he invested in consumer staples. Things everyone needs (or wants). If DDOG can live up to its hype (a very big if in my mind), perhaps they will become a “consumer staple” of future software applications.

I am still very wary of the seemingly high valuation of DDOG and the danger of what seems to be a “technology gone viral” as I stated in a recent post. Yet this gives me an interesting perspective to consider.

6 Likes

Wow, Thanks for the condensed version Saul. There’s a ton of meat in this report. As usual (for me) I’ll skip over the financial analysis, I’ll just assert it’s great. But bear with me while I make a few observations about the technology.

A very little historical perspective might be useful. Let me start with a few quotes:

Our customers can frictionlessly add-up new products or from this end-user interface, powered by a common data model. We win in the market for several reasons. One we are a truly integrated platform allowing us to solve our customers end-to-end problems and innovate rapidly. Two, we were built for the modern dynamic stack offering end-to-end visibility. Three, we are simple but not simplistic, easy to install with no professional services. And Four, we are designed for use in collaboration across development operations and business teams. I’ll explain the bolding in a minute.

From the Q&A: The reason for the low rate of use is that these legacy APMs are very, very, very heavy weight and they’re very expensive. It’s very hard actually to deploy them and get value out of them and it ends up being limited to a small set of extremely high value apps, for which you can be convinced to make an investment and get some ROI out of it.

It’s easy to skip over those comments without really understanding how incredibly important they are. In the old days (say maybe 10 years ago and longer) application performance monitoring (APM) was the dream of virtually every IT application group but virtually never implemented. The reason for this was in order to monitor the performance of an application, you had to instrument it. The instrumentation was fraught with difficulty, first, exactly what processes do you instrument in order to monitor the application as a whole while insuring that you are able to extract usable information in order to improve the performance, all the while being mindful of the Heisenberg paradox, the instrumentation itself had an adverse impact on performance. In case the term “instrument” is not clear, it means putting extra code in the application mostly in the form of counters and timers that separately reports out performance characteristics such as elapsed time to execute, number of loop iterations, number of database calls and so forth. Lightweight and frictionless - anything but, heavyweight, expensive, invasive and impactful are more fitting descriptive terms. It just didn’t get done and when it did get done it was usually in a test environment and then disabled or stripped out of the production code which meant you never truly got useful APM metrics on the code that was actually running. I have no understanding of how Datadog does what they do, or maybe it’s just a simpler task with more modern computing environments and languages - I don’t know. But lightweight, frictionless APM functionality has a huge market. Hence my bolding of the statement above.

Without going into as much of a detailed explanation I’ll simply assert that exactly the same thing holds true for network performance monitoring. And the ability to merge and analyze the data from these monitors is of immense value in problem avoidance in the first place and detection and localization when problems do arise. There’s a huge difference between knowing you have a performance problem and knowing where to look in order to repair it.

Skipping to another product that is probably not well understood would be log management. So first, let’s understand just what a log is. In short it is a chronological record of every transaction that occurs in a given bounded computing environment. It is almost impossible to read a log and get anything of value out of it as it is painfully detailed and seldom describes a contiguous string of transactions that led to a specific state in that every interrupt, every time trigger, every everything gets recorded. So what good is it? Well, the primary purpose of maintaining a log was that a computer could use it if things got hosed up to retrace and back out every action over a given period of time in order to restore the entire environment to some prior state. This was only done in the most extreme situations because you not only backed out whatever gave rise to the problem, but you also backed out all those transactions and state changes you wished to retain. It is entirely non-selective.

On the other hand, if you could parse a log and sort it into strings of related activities it could be of tremendous value with respect to understanding the factors that are needed for load balancing, resource allocation and even infrastructure purchasing decisions. There’s lots of valuable information in the logs, but it’s not useful in its native form.

One final observation, and this one is speculative in nature, sort of hidden in the weeds. But the CEO opened his remarks with We exist so that our customers can understand everything that happens in their technology stack. And later there are references to we announced enhancements to our machine learning and AI capabilities. Watchdog, Metric Coalition, Trace Outliers, etc. When I read that it struck me as leading to a powerful security capability. Almost all security breaches have at least one thing in common, that is that they rely on anomalous, unauthorized behavior going undetected. Going back to the Target breach for example. The invaders set up a staging database within Target’s own technology stack that they sued to gather and organize the hijacked information before shipping it off to their own servers. This is most often how breaches are planned. It’s more secure (for the hijackers) to stage the information on premises than to ship it off one transaction at a time. Making the entire stack transparent and setting off alarms for unusual activity could easily thwart many security attacks.

36 Likes

Wow, Thanks for the condensed version Saul. There’s a ton of meat in this report.

Yes Brittlerock, and thanks for the kind words, there was indeed a lot of useful information for me in the conference call.
Best

1 Like

Hi Saul,

Thank you for this amazing summary of DDOG’s earnings report! I had tentatively dipped my toe in the water with a few shares a couple weeks ago when they really dropped a bit for no reason. But this report gives me a significant boost in confidence for their potential!

Two things stand out in this report for me:

• In October, we announced we are currently “In Process” on the FedRAMP Marketplace, initiating the certification process. Achieving FedRAMP authorization will expand our addressable market, by allowing U.S. federal government departments and agencies to adopt and use our cloud platform.

Having worked at a start-up which had spent a considerable amount of time, effort, and money chasing FedRAMP certification, it’s important to note just how difficult this rating can be! My former employer spent more than 7 YEARS chasing FedRAMP certification. It is a long, tedious, and arduous process. And even now, after 9 years, they still don’t have FedRAMP certification!

So, why have they continued to pour so much time and money into it? Because they believe the amount of future revenue they’ll gain as a result of that one certification will make it all worth it! You can not sell into many of the government agencies without this certification! And once you have it, you are often THE ONLY ONE in your category with that certification. This means you have a complete monopoly on all sales into the Federal government for your product category!

It doesn’t matter how good, or even how many competitors you have in the overall market. If you’re the only one with FedRAMP certification, the vast majority of government agencies CAN NOT buy from your competition even if they want to!

Which leads to the second quote which stands out:

We believe we are at the very early stages of an existential market opportunity,

If they are close to getting the FedRAMP certification, then this is a very true statement. But, more importantly, this statement is indicative of the ideas others have stated. We have not even begun to see the expansion of a SaaS-based world yet. In my previous company my eyes were opened to what SaaS and cloud based applications will look like. The future of software development is completely inverted in the cloud. Everything is possible, and everything is completely decentralized. Gone are the days of a fixed environment in fixed data center. Gone are the days of having an IT staff “own” the data centers and networks and monitoring, etc. In the future, all this will be owned by the developers themselves.

Monitoring is an essential component of running a software platform. And everything that can be monitored and measured should be, otherwise, how do you know where problems lie, or improvements can be made? However, no one likes monitoring. It’s tedious, it’s cumbersome, and it’s HARD. And where do you draw the lines? DDOG has solved all of that and provided a flexible, customizable, and robust platform to monitor everything and anything you could ever want for cloud-based SaaS platforms. They’ve gone viral (as someone earlier stated) because they’re that good. You could choose lots of alternatives, but none of them are as comprehensive, or as flexible, or even as cheap once you consider the amount of effort required to cobble all the pieces together and the difficulty of integrating all those pieces into your solution.

And since DDOG is essentially a first-mover in this space, AND we haven’t even begun to see cloud-based SaaS take off yet, I believe they are right; we are sitting at the threshold of an existential market opportunity!

The big question now is: Can they execute and maintain their lead position in order to exploit that opportunity! Only time will tell!

Anyway, thanks a lot for this summary, I really appreciate it!

Paul

14 Likes

If you’re the only one with FedRAMP certification, the vast majority of government agencies CAN NOT buy from your competition even if they want to!

Before we get too excited, most of the main competitors to DataDog. New Relic, Splunk, and Elastic to name a few are fully FedRamp certified and have been for some time. They get a lot of Federal Government business.

I was happy to hear that Datadog was at least getting started with this designation, but a little disappointed that they weren’t a little further along than this at this point.

Darth

15 Likes

most of the main competitors to DataDog. New Relic, Splunk, and Elastic to name a few are fully FedRamp certified and have been for some time. They get a lot of Federal Government business.

Darth,

Can you point to where you found this information. My cursory search has yielded nothing to that effect.

Splunk apparently went through the GSA PMO for certification and received the Moderate Impact Level Authorization, however they’re the only ones.

New Relic (as of 1/9/2018) is “nearing FedRAMP ATO” certification. I can find nothing on their site about having received the ATO, or at what level they are seeking authorization.

Elastic’s website merely says “Coming Soon”, so presumably they’re in the same boat as New Relic, and still waiting…

Also, apparently the sponsoring agency for FedRAMP also matters. For example, DOE supposedly has the toughest requirements for their vendors of all the agencies, therefore, if you meet the DOE’s standards for FedRAMP certification, you can sell into any agency because you already meet the most exacting requirements (I learned this being involved in the FedRAMP audits for my previous employer). I believe this is what is known as “High Impact Level” authorization. Whereas, Splunk, as noted above, comformed to GSA’s “Moderate Impact Level” authorization. Which means Splunk could sell into any agency at that level but not into agencies like DOE, CIA, FBI, Treasury, etc. who require High Impact Level Authorization.

As I noted earlier, it can take YEARS to get FedRAMP ATO certification. So, the fact that DDOG is only now “initiating” the process is rather disappointing. My hopes were a lot higher based on the earning summary where it was mentioned. Investigation now leads me to believe we’ll be waiting for this for many years to come…


Paul

9 Likes

Hey Paul,

My apologies to everyone. It appears I was mistaken. Must have gotten confused in the language.

You can check the FedRamp marketplace for particulars.

https://marketplace.fedramp.gov/#/products

Splunk - Authorized
New Relic - In progress
Datadog- In Progress
Elastic - nada

I guess you can get a lot of Fed business without it.

Darth

5 Likes

I recall reading Docu and ZS are FedRamp authorized but it has not really taken moved the needle. Couple quarters ago Docu in fact said it will take a long time before you will see much rev.

DDOG seems priced for turbo charged growth. I recall when ZS accelerated to 65% its valuation took off. But with some softer commentary last CC things have really fallen off a cliff. Not saying this will happen with DDOG of course. It is definitely on my watch list. ZM on the other hand is slowing rev growth at about 8% every Q. 2 years ago it was 149% for the year, last year it was 119%, this year will be 92%. If they keep at that rate a year from now they will be <70%.

2 Likes

DDOG seems priced for turbo charged growth. I recall when ZS accelerated to 65% its valuation took off. But with some softer commentary last CC things have really fallen off a cliff. Not saying this will happen with DDOG of course. It is definitely on my watch list.

Hi Tex Mex,

Zscaler says that they have to sell to C-Level people who can authorize changing a big company’s entire security orientation, and that this now means slower sales.

Datadog says that they tell their salespeople not to try for the whole company but make little, no-friction sales to get into the enterprise, and it will spread virally within the enterprise. It’s so easy to adopt that they have 100% subscription revenue and no service and hand-holding revenue. They don’t need it.

They are entirely different animals.

Saul

11 Likes

Datadog says that they tell their salespeople not to try for the whole company but make little, no-friction sales to get into the enterprise, and it will spread virally within the enterprise. It’s so easy to adopt that they have 100% subscription revenue and no service and hand-holding revenue. They don’t need it.

As Bert wrote about DDOG in Sept:

It most recently reported a dollar based net retention rate of 146%, which is about as high as it gets.

Compare to Zscaler which was 118% the last two quarters, because they have to try to sell the whole company up front.

Saul

3 Likes

Yes, good point about ability to sell Datadog lot more easily than ZS. I suppose my biggest concern with the trio of super growers (Ddog, Zm, Crwd) is that there does not seem to be a whole lot of past history of companies with sustainable 80%+ growth. CRM et al. quickly slowed. So, if these companies slow down into the 60%s over the next 12 months the stock price will hardly move and may even drop.

1 Like

The benefits of Fed on ramp go beyond the ability to sell into those agencies that require it. If you achieve on ramp it basically means you are running a really tight ship. For some unique insight see the podcast below.

https://podcasts.apple.com/us/podcast/a16z-podcast/id8428187…

1 Like