Intelligent automation is rapidly transforming the global economy, delivering momentous gains to enterprises that adopt it at scale. One recent article by McKinsey revealed that some organizations have been able to automate 50 to 70 percent of their workflows, generating ROI that reaches into the triple-digits. In addition to cutting costs, intelligent automation can also deliver precision, speed and enhanced customer experience.
In order for organizations to enjoy the full value of intelligent automation, IT leaders must be willing to take a guiding role. Unfortunately, many IT executives find this challenging, whether due to the increased complexity of IT processes, lack of understanding and/or clarity, inconsistent or fragmented tools that hinder scaling, or the misconception that intelligent automation cannot be adopted without major re-engineering of existing processes.
How can these challenges be overcome? And how can IT leaders succeed in their automation initiatives? The answer to these questions lies in the following four key steps along the intelligent automation journey.
Step 1: Evaluate the high-level potential value
The first step in becoming an intelligent automation leader starts with the development of a clear business case. This involves assessing the potential high-level value of the company’s main IT activities. Some examples of what these areas of value might look like include:
Incident Response – A significant number of IT incidents are initiated through support desk requests. These typically result in tickets being created and assigned to Level 1 support agents. While these are the obvious candidates for automation, the portion of tickets that are escalated to specialized L2 and L3 agents are also ripe for the picking, thanks to the advanced technology behind intelligent automation. And since these activities are generally well-documented, categorizing and prioritizing them by automation potential should be relatively straightforward.
Planned Activities – In addition to the one-offs and unexpected support tickets that crop up, IT is also responsible for performing a number of planned activities on a regular basis. These activities typically include things like backups, upgrades and patching. They may also involve more complex security audits. The amount of time and resources required to perform these duties can collectively add up to around 20 percent of the IT budget. Calculating this figure can help determine the potential savings intelligent automation can deliver.
Introducing New Applications – From a business perspective, this activity is often viewed as the one that produces the most significant value. It can also account for an additional 20 to 40 percent of the time and resources put forth by IT. These activities are not exclusive of application development, either. They also include such tasks as testing and hosting. This places increasing demand on both the application team as well as the infrastructure group.
Step 2: Dig deeper to identify which specific use cases are best suited for intelligent automation.
Determining how to effectively implement intelligent automation requires a deep dive to uncover the root causes of issues. It may also involve the untangling of complex systems and the development of an accurate picture of how to leverage automation to extract the greatest value. In other words, the process is a complicated one and requires a certain degree of commitment. Let’s take the three potential use cases above as an example.
Automating IR begins with identifying which incidents are the best candidates, which can be challenging. The goal should always be digging deep enough to uncover the “why” of documented incidents. Without this information, efforts are futile. Text-mining can help by automatically reading ticket descriptions and extracting the necessary insights to sort them into three categories:
- Requires machine learning
- Highly cognitive/manual
This analysis should leave you with a prioritized list of incidents to automate and the type of automation best suited for the job.
Most enterprise-grade IT departments rely on industry-standard tools to manage their infrastructures. Unfortunately, due to factors such as advanced customization, adjustments due to mergers and specific user requirements, managing these systems often requires a significant amount of manual effort, diminishing the overall value.
For instance, despite the widespread adoption of infrastructure and application monitoring tools, support teams are often unable to respond effectively to the logs being generated, either because there are too many of them or because of the many reasons why they are being generated in the first place. As a result, IT leaders are often unclear on how to approach intelligent automation implementation.
In situations such as this, machine learning technology can be “trained’ to identify the reasons behind alerts and then either recommend or autonomously make better decisions on which action to take. This eliminates much of the complexity for the IT team.
Introducing New Applications
Many IT executives fall into the trap of focusing solely on the reduction of manual labor. As a result, they fail to see and achieve the full value potential of intelligent automation. Faster and more accurate delivery of applications requires the development and design of a new operating model, with an emphasis on DevOps and agile.
Reviewing this entire process to gain an understanding of how to make the most use of this new operating model can result in entirely new approaches to work. Intelligent automation can facilitate some of these new ways of working. For instance, automating the testing process will enable applications teams to iterate more quickly. Likewise, developing a self-service model for things like automated server provisioning allows the operations team to become more responsive. The list goes on.
Step 3: Execute your proof of concept
In order to demonstrate the true value and validate your case for intelligent automation, the next critical step is executing a proof of concept. A great place to start with this is incident management. Organizations that have successfully deployed intelligent automation for incident management have been able to achieve substantial cost savings in a relatively short period of time.
Thankfully, there are many different incidents that can quickly and easily be automated to support your proof of concept, including such tasks as password resets and employee onboarding. In its most basic form, a proof of concept requires the following:
- Collaboration with subject matter experts to identify where automation can best be applied and understand all the steps and systems involved in a particular process or workflow.
- Careful selection of an intelligent automation platform. Look specifically for products that can be integrated with existing systems and applications and offers pre-packaged, no-code options. (This will enable rapid adoption and time-to-value.)
- Obtaining necessary IT and overall business approvals with regard to regulatory constraints, security guidelines and access limitations.
- Ongoing testing and monitoring to capture results and document value
This phase is also an ideal time to consider building stronger internal intelligent automation capabilities; for example, developing a team to spearhead a future automation center of excellence (CoE). This team will ultimately become the foundation and engine that drives all IPA initiatives.
Step 4: Build intelligent automation capabilities to scale
Achieving the full benefits of intelligent automation requires the development and nurturing of certain skills and capabilities, in addition to rolling out an entirely new company-wide culture. This is essential as successful adoption of IPA requires that automation become embedded into the very heart of the organization itself. There are plenty of ways to accomplish this, but generally speaking, companies that have been successful have done the following three things:
Build on success to expand into new areas of IT (and beyond).
Once the basic tasks and workflows have been automated, it’s time to move on to more advanced level-2 and level-3 activities. The IT team should be expanding beyond incidents to begin leveraging the AI and machine learning technologies to assist with things like analytics and decision support. The goal is to eventually roll out intelligent automation to as many routine and complex processes as possible.
Spread the word.
With a strong foundation of capabilities and experience, IT leaders can begin to position themselves as subject matter experts for the rest of the organization. This process involves continued outreach, such as connecting with other leaders across the enterprise to advise them of the specific benefits IPA can have for them. This outreach also provides the opportunity to identify additional areas where automation might be beneficial.
Explore the advanced elements of intelligent automation.
While the majority of organizations have thus far only focused primarily on simple process automation, the future belongs to those with an eye toward artificial intelligence and cognitive learning. These solutions are already making an impact on companies with forward-thinking leaders. The best way to break into this arena is to start working on small AI initiatives. From there, just like basic automation, you can continue to build, expand and grow.
Intelligent automation is maturing rapidly and quickly becoming a core component of the IT landscape. IT professionals who recognize the importance and understand how to develop their automation capabilities have the potential to become respected leaders in the process – a title that will serve them well throughout their careers.