Dynatrace CEO Rick McConnell’s first job out of college was working as a financial engineer at an investment bank, programming IBM mainframes in FORTRAN. He went on to hold senior leadership roles outside IT like CFO and CEO (most recently serving as President at Akamai, prior to joining Dynatrace in 2021) but sitting down to talk in person in Las Vegas this month, he tells The Stack he still does “really love the technical aspect of things.”
Dynatrace’s métier is automated discovery and mapping of applications and their infrastructure components. That helps users to find out where things have gone wrong and how to fix them. Its customers do not, necessarily, really love the technical aspect of things however. For most their IT stack is utterly business-critical; it’s also increasingly dizzyingly complex, and the evolution of this sprawling infrastructure is often a pain in the back-end.
Give The Stack a follow on LinkedIn here
(One customer, TSB’s COO Suresh Viswanathan, gave a good example of this to The Stack in a 2021 interview. As he put it: “If you call ‘0800 TSB’ today, before we ask for your mother’s maiden name, the call lands with a managed service from BT; the identification is then a managed service from Fiserv; the transaction works its way through to Microsoft Dynamics implemented in Azure; then it goes to the banking platform, which runs in an IBM plant to render the screen before the agent says: ‘Hey, what’s your mother’s maiden name?’ That’s the architecture we run. On a bad day, when I’m not able to say ‘what’s your mother’s maiden name?’ — because the problem could be with BT, Fiserv, Microsoft, IBM — you could be left saying ‘hell, what happened?’”)
Dynatrace has grown to over $1 billion in revenues (hitting that milestone this year) by helping organisations understand such infrastructure bottlenecks or blindspots. The software intelligence firm provides application and infrastructure performance monitoring and observability that spans traditional hosts, virtual machines (VMs), containers, networks, and events.
But Dynatrace, which runs the majority of its R&D out of Linz, Austria, where CTO Bernd Greifeneder founded the company, is increasingly branching out, including into a log management and analytics world more typically populated by security companies. It describes this as “a market that is ripe for disruption” and in October 2022 launched a new Dynatrace product called “Grail” with the aim of being the disruptor in a $50 billion dollar market that McConnell describes as full mostly of “DIY dashboards.”
What is Dynatrace’s Grail anyway?
Dynatrace’s press release at the time sailed past The Stack with its jargon-raddled white noise about “unifying observability data as well as security and business data from cloud-native and multi-cloud environments, retaining its context, and delivering instant, precise, and cost-efficient AI-powered answers and automation.”
Grail is, however, an interesting proposition that is now generally available.
Grail is a data “lakehouse” (a happy medium between the original Big Data storage “data warehouse” and the more analytics-friendly “data lake”) that provides schema-less (flexible) ingestion of various data types, including logs, metrics, traces, user data, business data, and topology data; all things that can give you a birds-eye view of what’s going on in your infrastructure, spanning performance, trouble-shooting, and security.
The idea is full-stack observability with automated Root Cause Analysis (RCA) support, even if an application issue is tied to a microservice on the back end, and be able to do it in as close to real-time as possible without having to “rehydrate” data – reconstitute compressed files, or pull them out of cold storage for analytics.
McConnell says: “We believe it's [the logging sector] is ripe for disruption, because the analytical capabilities that exist, especially with log management are not real time; increasingly, organisations want to be able to run real time analytics on logs. Most data and logs have to be indexed. Then depending on the queries that you're running, it has to be reindexed. So complex queries can become very lengthy. With Grail we can provide – by not doing any rehydration or reindexing, based on our graphic architecture that we've deployed – near real time analytics on logs. That provides much much better performance that we believe because we had no innovator’s dilemma around something that existed before – gives us the ability to leapfrog that market.”
Even prior to Grail’s release, Dynatrace was increasingly supporting customers with security insight, he notes: “The example that I would use is Log4J, which hit in December of last year. All of a sudden, you’ve got a set of vulnerable libraries. Because we sit deeply in your infrastructure, we could tell customers within 15 minutes exactly where they were using all the Log4J library calls and the number of calls in each of those areas, so you knew not only all the places you were using Log4J, but what order should you be deploying patches against it.
“This is where the observability data and the observability oversight that you have, makes the application security capabilities work better, and vice versa. It really does provide an invaluable deep interconnection.”
Dynatrace CEO Rick McConnell on profitability, serverless, and cycling
Dynatrace’s financials are robustly healthy in a challenging market. ARR has hit $1.06 billion; that’s up 33% year-over-year for Q2 of its fiscal 2023 (ending September 30, 2022). It added 164 new customers in the same period and saw a net expansion rate of above 120% for the eighteenth consecutive quarter. Unusually for so many software companies grown fat in an era of easy money, it’s profitable (on a GAAP basis, since 2020).
Dynatrace CEO McConnell says the ability to demonstrate ROI for CIOs helps: “In our case, it's a combination of what is the value to your engineering teams, being able to be much more efficient about correcting issues. But also, the value of having an IT operations centre work much more efficiently. It isn't just about data and data on top of data. It is about delivering very precise answers based on contextual data types, logs, traces, metrics, etc, real user data, so that you can isolate and understand what does good look like? What do normal ops look like?”
There’s no shortage of software companies that have been stung by a rapid move away from on-premises infrastructure and software to a world of public cloud, microservices, containerised applications.
Dynatrace’s consistent revenue growth and net expansion has been helped immensely by the company being able to straddle a multiplicity of infrastructures thanks to its R&D team (which gets allocated around 15% of the company’s annual revenues) and leadership keeping a close ear to the ground with customers, he says.
“We spend a lot of time with customers understanding where they’re heading. For example serverless, OpenTelemetry; these kinds of elements where they’re saying ‘wait a minute, you can’t put your agent on these machines because it doesn’t work the same way’ – but we began around four years ago anticipating a move to the cloud that was going to be that was going to be persistent, and pervasive and did a complete re-platforming four years ago, to be prepared for a competitive environment” adds Dynatrace’s CEO.
Dynatrace says that this has helped it get ahead of this evolving world with its PurePath code-level analysis technology supporting out-of-the-box insight into things like Istio/Envoy service mesh communications as well as Kubernetes services from all of the hyperscalers. And whilst companies and indeed engineering teams might be building tidy self-serve capabilities (whether that’s using native observability capabilities of the platforms they’re using or things like OpenTelemetry, OpenTracing, OpenMonitor, OpenCensus, etc.) their risk is that they end up with fragmented insights into the different parts that make up a modern application, he notes.
He gives the example of a new customer in the oil and gas space (declining to name it) that is running over 4,000 applications across multiple cloud providers and recognised the need for a unified observability platform that even an “army of people” in network operation centres couldn’t oversee and triage completely.
“They walked me into an Operation Centre in Houston and said ‘Rick, this is what you need to fix for us.’ That was a really good depiction of what the concerns are out there: That my existing processes are not going to work with my current observability solutions and this is being exacerbated by the move to cloud, which is causing the number of apps, the amount of infrastructure, and the data flows to expand uncontrollably.”
(Talking of oil and gas, however far removed, if there is one area in which the company lags a little, it’s ESG. The NYSE-listed company has yet to understand and track its emissions, although it is currently “undergoing a greenhouse gas (GHG) emissions accounting analysis to inventory baseline environmental data and understand our carbon footprint” and plans to publish its first ESG report in 2023. That will, an initial paper suggests, include “understanding and addressing our greenhouse gas footprint and climate-related risk across assets and activities with operational control (Scope 1 and 2), as well as our supply chain impact (Scope 3), including the energy consumption of Cloud services" -- good luck!)
As CEO, McConnell’s getting this particular interview, graciously, in a noisy temporary booth at AWS's re:Invent jamboree in Las Vegas, but it would not be unkind to note that a lot of the innovative work done at Dynatrace and indeed the work to ensure it could continue to play in serverless-first world happened before his arrival a year ago. What does he bring to the table as a CEO and what is he trying to achieve?
“In my view and based on my experience what gets you to $1 billion isn't what gets you to $2, $3, $4, $5 billion. So I come in just under a billion. At Akamai we managed to take Akamai from $1 billion to $3.5 billion during my 10 years there; at Cisco we took it from under a billion to $5 billion in terms of communications.
He adds: “Certainly one thing that I've learned and really tried to incubate is this notion of empowerment. You’ve got to trust your team. You've got to bring in the right people, and empower them to win. The other thing I really spend a lot of time culturally, trying to figure out how I can have people in Dynatrace say, ‘I can't imagine working anywhere else, I love this place! I love what we're doing, I aspire to help deliver the vision of software working perfectly. I love the people that I'm engaged with, I’m paid fairly, the benefits are good.’”