IT automation is a challenging product category to evaluate. For instance, for every big technology vendor that has a suite of tools, there are multiple massively popular point solutions that do their task remarkably well. This high degree of fragmentation often means that end users select solutions based on cross-product integration as much as functionality or cost.
This approach often works, because enterprises typically like the ability to assemble their own bespoke tooling stacks. That said, a best-of-breed approach, whether integrated or not, does raise its own challenges. For instance, when a problem arises, different interpretations of the issue coming from different automation tools can lead to slower diagnosis—and, in the worst case, ignite the blame game between teams.
But in the bigger picture, the whole category is changing as vendor consolidation and AI integration begin to alter the IT automation landscape. I predict that the IT automation market will look quite different two years from now—and I believe that’s a good thing for enterprise customers.
The Downsides Of The Status Quo In IT Automation
Let’s start with the challenges that IT automation has had to face:
- Multiple versions of the truth — Automation tools are often targeted at a specific IT role such as DevOps or a specific layer of the stack such as containers. So, we often see a DevOps person using one toolset and an SRE using something else. This can create challenges when a crisis arises. A single event can kick off multiple alarms that may trigger an uncoordinated response. Additionally, it can be a challenge for teams to develop the right path to achieve issue resolution because of different interpretations of the problem. So while you may be getting great visibility and analytics, you may not be getting great diagnosis and remediation.
- Event-driven approaches — IT automation tools are often event-driven. Events can be something routine like initiating a software update, or ad hoc, such as rebooting a hung server when a specific trigger is tripped. These types of robotic actions can be very helpful in enabling IT resources to better scale their efforts. However, this approach can also present the issue of programmability versus exception handling. Developing event-driven workflows can be hard to program in heterogeneous environments, given the number of exceptions that can be generated. Also, it’s a challenge to create enough workflows to handle all of the different possibilities. And none of these things are great for black-swan events such as the recent Crowdstrike BSOD incident.
- Tech-centricity versus business-centricity — IT automation tools historically have come from the perspective of reducing IT costs by increasing the capacity of existing IT resources. And while that has indeed happened, there are other aspects of operating IT such as forecasting capacity, optimizing cloud deployments and enforcing business compliance rules that are not automated. We are starting to see IT automation vendors take some of these elements into account. Some of the more visionary vendors are actually leveraging ideas from supply chain management to provide a more holistic and business-driven approach to IT. (More on that in a minute.) But there is a long way to go on this.
The Upsides Of The New Wave Of IT Automation
As I touched on earlier, there are many changes reshaping the IT automation market, and I believe they will end up being for the best over the next couple of years. Here are some of the positive trends driving a behind-the-scenes revolution in how enterprises will manage almost every aspect of delivering IT services in the years to come.
- Market consolidation — Large tech vendors have been making big moves in the IT automation space. For example, while many think about Broadcom’s acquisition of VMware as a virtualization play, VMware’s IT automation stack was also quite robust and was able to manage a lot more than VMs for a long time, so the IT automation aspect of that deal should not be overlooked. Meanwhile, IBM has spent the last few years building an IT automation stack based upon organic development for solutions such as IBM Concert as well as multiple acquisitions including Apptio, Turbonomic, Instana and now HashiCorp. These big vendors are all intent on providing a more unified and coordinated approach to IT automation. And it’s not just the big players: we are also seeing smaller companies in the space adding assets via acquisition to improve their stance in security or AI.
- AI and observability — Like most technologies, IT automation is going to get a big boost from AI. In the case of IT automation, that boost comes from both machine learning and generative AI. Machine learning will greatly enhance observability and create predictive patterns to set triggers and automatically create the right workflows. Also, we are seeing GenAI used to help relay suggestions to IT for resolving issues as they happen. But what is most important is that the underlying data models driving the AI will also help unify the disparate silos of information across enterprises. Take Dynatrace: while it’s a good example of a company that has touted the effectiveness of a single IT automation platform for years, it has also introduced its own AI solution to enhance how it predicts issues and builds workflows to mitigate them. Long term, you will also see AI used more actively in the optimization of the application. For example, there may be a performance advantage in changing to a different storage model or database. A well-trained AI may keep an eye on an enterprise’s approved vendors to find the most effective technology—and even the best business deal—as the market changes.
- A supply-chain-based mentality — IT automation’s evolution is starting to resemble how companies have evolved to manage physical supply chains. One of the biggest similarities is that supply chain management is no longer an isolated function but a cross-functional team effort. When a new application is planned, it’s not just the technology requirements that are important for selecting components and services; business factors such as supportability and budgeting also need to come to the fore. And given the highly distributed nature of application services and infrastructure, regular reviews are mandatory, not just from an operational viewpoint but also from the standpoints of human capacity, costing and compliance. Many smart IT operations are following these practices now, but the tooling has not always caught up. Two companies that are really starting to dig into this supply-chain concept are IBM and JFrog. While their products are targeted at individual practitioners, each product portfolio works holistically to address the end-to-end application lifecycle. These solutions are a good analogue to how physical supply chains work in the real world.
Further Evolution Required
The big question remains whether all of this change will actually live up to the potential. I believe it will, but there are a couple of caveats that need to be addressed. First, the ROI model will need to evolve. These solutions can become expensive, and if the value proposition is simply IT cost reduction, it may not be enough. Other considerations such as the time to deliver a compliance audit or the value of higher-quality software—in other words, the delivery of business outcomes—will need to be considered.
Complexity and overlap must also be addressed. Larger vendors have quickly updated their IT automation stacks via acquisition, so there will need to be some refactoring efforts. Also, third-party integrations will still be necessary. As I said at the beginning of this piece, there are some tools out there that are best-of-breed for what they do, and customers using them will likely be reluctant to move away from them. Future winners in this market will need to be able to showcase both innovation and integration.
That said, the potential for more holistic, predictable and automated IT operations has not been this high since the pre-cloud days of large monolithic systems management suites. This is good news.
