A case of 'the three kings’ from Dynatrace?

Martin Banks Profile picture for user mbanks May 2, 2018
Dynatrace CEO John Van Siclen argues the move to open systems allows management systems to complement each other to provide the complex management environments real-time cloud services require.

Three Kings
Over the next five to 10 years, most, if not all businesses will move from on premise into the cloud, which will bring about a serious shift in the brands that most of those businesses will turn to help them make it work.

This is the view of John Van Siclen, the CEO of Dynatrace, which has turned Application Performance Management into an AI-based cloud environment monitoring and business operations (Biz/Ops) management platform that has generated a growth rate currently running at some 800% year on year.

He is expecting to see a new cloud-oriented hegemony emerge built around a triad of Dynatrace, Splunk for security management and ServiceNow for operations management. The keys linking them together are openness, the data they generate, and the ways in which they can exploit it when it is shared.

This movement is emerging so fast it doesn’t really have a name yet. Back in February I suggested that Autonomous Business Outcome Management might get close and Van Siclen acknowledges that `business outcomes ‘ are now the real goal of the users now:

It's interesting that you should bring that up because what I've found is that people are saying about the cloud: 'well that actually works, I'm going to rebuild my entire IT infrastructure around dynamic enterprise web-scale cloud.' But they think about building an agile platform where monitoring is part of the fabric. They haven't yet flipped it to, 'Well, if I know everything that's moving from the tap/click/swipe of every user, all the way through the entire infrastructure, don't I have a business outcome-driven intelligent engine?' and the answer is yes.

There's no question that the analytic engine we've built and analytics in general are going to define the next generation of software success. If you don't have very high degrees of automation and intelligence, why bother? And that's actually something interesting with these cloud architects; they actually get the fact that automation and intelligence are two critical components that they need.

'Muscular’ use cases

As examples of the scale at which Dynatrace is now operating, he cites a UK bank which is planning on putting two applications, one with 30,000 containers and the other with 40,000 on a platform, in the next six months. He talked of another that is going to completely change the compute infrastructure from eleven data centres, down to three running a completely cloud-based, containerized platform between now and 2022. A third is changing the way its entire operation works:

Instead of doing the big bang, the CIO is taking a piece at a time, and what he's doing is setting up user experience at the edge because he wants to know how every user is doing. This is because he's going to be swapping out components in-between, so he wants to make sure that, as he's swapping components and upgrading, everybody is happy. They're a very consumer-driven sort of bank.

Scaling for such environments, and their future growth, is likely to be helped by the fact that Dynatrace can be tiered, allowing users to set up Dynatrace-managed suites of Dynatrace environments.

The company’s rate of growth, coupled with its potential for a far wider and deeper penetration into the way businesses manage their operations in the more complex, real time world of cloud services, is likely to make it an acquisition target for one or more of the major business systems vendors, a possibility that Van Siclen is well aware of, to the point where it holds a corner of his mind.

Dynatrace is currently owned by the private equity management company, Thoma Bravo, and as he observes, private equity guys are usually in the business of a transaction somewhere:

Our approach is, whether it's strategic who comes in, or whether it's another private equity company that wants to keep it private and ride the momentum, or whether it's an IPO, we'll figure that out over the next couple of years. There won't be more than a couple of years where one of those will happen.

He also thinks that, at last, stock market analysts have caught on the longer term potential of the annuity business model on which cloud services are predicated. This suggests he feels an IPO or similar would now be met with a far more favourable stock market response:

The market has got quite sophisticated that the software as a service subscription is the right kind because it creates predictability in growth. And so, everybody watched 'the Salesforce movie' with a little bit of scepticism early, over the last five years it has become a question of, 'where do we find more like that?'. We're not just going through a tech refresh, we're going through a business model refresh at the same time because with Dynatrace, two thirds of our business is on subscription. So it's starting to create a flywheel and more and more and more is being bought as a subscription.

Open sesame

The key to this development, he feels, is the not just the technology developments within the Dynatrace tools and services portfolio, but also the conscious decision to open it all up as a platform that users can access in ways appropriate to their needs:

The minute you start moving from a product to a platform and you open up and enable APIs everywhere, people will do amazing things. And you can look at some of the companies that are on the path ahead of us. Three years ago, ServiceNow announced they were going to be a platform and opened themselves up. Splunk has been there a couple of years ahead of that, and you see their success. Same kind of thing for us.

He sees the possibility of a complementary relationship between the three of them, though only a higher, more abstracted level, however. Closer to the technology there are differences that would probably get in the way. For example, he stated that Splunk uses data scientists for log analysis while Dynatrace used AI to couple log data with other data sources. And while ServiceNow has some interesting and releVant data sources, it is a platform that sits outside the cloud. He does not see this as not a good place to manage processes inside a real time cloud environment:

So I think we are complimentary, because even though we live inside the cloud and they live outside the cloud, there's a lot of data and information that can go back and forth to make both systems better, smarter and more effective.

An example of this would be in ServiceNow’s access to change management data. Dynatrace can pinpoint the root cause of a problem in a very large scale cloud automatically, but can’t tell whether it was caused by a change or not. But if there is an API that could move the change data into the Dynatrace AI engine it would be possible to identify that a simple roll-back of the change could rapidly solve a problem. This could provide a very precise self-healing capability.

He sees Splunk's primary business now moving very much into the security arena, which makes it very much a complement to the other two.

This does seem to be a very able demonstration of how `openness’, once the bete noir of all proprietary-oriented IT vendors, will benefit not only users, but the vendor community itself, as Van Siclen identifies:

So here is one of the things that I see right now across the market. The CTOs and CIOs don't want a lock in, so they pull back from AWS and insert themselves into a platform in between. This is why RedHat, OpenShift and the Pivotal cloud foundry guys are exploding. We think we're exploding. They're exploding in the same way. This is because people want that layer of portability so they can do hybrid or public cloud, and in public cloud they can do multi-public and all of that, wherever and whenever they want to run it. So if they are not happy with the deal they got, they can just move the workload, since they are not locked anywhere.

But that does require a broadly focused set of complementary, pan-environment management tools that take away the base decision of, "Can I move at all?' and replaces it with something more positive, such as 'Is a move to cloud environment X better or worse for my business outcomes?'.

My take

There is an old and well-known saying, that 'the whole is worth more than the sum of the parts’, and that is arguably true of all the best developments in any field. But when it comes to exploiting IT to get the best possible business outcomes, this is now the time that it really comes into play. Managing complex, fast changing, real time cloud environments that have more agility than the sum of an Olympic squad, is not a task for a single vendor – this is now very much a team sport. And openness is the only way to get there.

There will no doubt be more players coming in to contribute to this type of management environment, filling in corners here and there, and because it is open if it works they will find a welcome. And because it is open Van Siclen's’ observations should not be taken as any sign of formal partnership between Dynatrace, ServiceNow and Splunk. Such a thing would probably get in the way, and that whole idea is now no longer of great relevance. Now the only questions for users will be, `Will X work with Y to make Z? Because it is Z that I need for my business’.

A grey colored placeholder image