A path towards end-to-end automation
Customer service bots, digital payroll workflows, fleet management and lest we forget the clever programmer who famously automated his own job -- the uses for and adoption of automation continues to expand. In fact, the automation market is expected to reach $136.5 billion by 2026 with a five-year compound annual growth rate (CAGR) of 3.7 percent according to Industry ARC’s Automation Market Report.
But as growth and excitement build, confusion in the market also abounds. After all, the word "automation" is so overused and broad that it’s no longer descriptive.
Let’s clear up the differences between task-based and end-to-end automation, explore the best use cases for each and dive into how IT professionals can move towards end-to-end automation for added value.
Task-based versus end-to-end automation
Task-based automation is a common solution that uses software to reduce the manual labor involved in repetitive, low complexity work. This automation usually tackles isolated tasks like opening trouble tickets. Although task-based automation streamlines processes and increases employees’ productivity, human operators are still required to either start or end the workflow, and it has limited use cases. In many IT organizations, there could be hundreds of task-based "scripts" written and running, most without management even knowing. Script "sprawl" can be a real problem.
End-to-end automation handles longer, more dynamic workflows by stringing together multiple actions with logic and contextual data. Unlike task-based automation, end-to-end automation seamlessly coordinates with human operators and integrates different systems and tools into the same workflows. This type of automation provides advanced functionality and can handle repeatable consumer-facing tasks like customer service queries.
End-to-end automation requires orchestration, which is not unlike the literal definition of "orchestra." In an orchestra, different types of instruments come together to create an entire piece of music. And for this process to work, the musicians need a conductor, or an orchestrator, to guide all of the moving parts and ensure smooth transitions. The same is true in end-to-end automation. An orchestration platform enables end-to-end automation by translating and standardizing data and connecting disparate tools and systems.
Enabling end-to-end automation
End-to-end automation can be as hard to achieve as it is valuable. IT teams need to understand the entire workflow before automating it. And this automation requires redesigning workflows, analyzing the data that comes out of every automated task and making decisions based on the contextualized information from each step.
IT leaders looking to implement end-to-end automation should follow these critical steps to ensure success:
- Build the strategy. What business problem are you trying to solve? What is the potential impact?
- Decide which process to automate. Which automated process could solve the business problem?
- Map out the process and workflow. What are the data sources? How will the data be standardized? And where are the decision points?
- Architect the process. How does the automated system access the data sources -- through APIs or through a different interface? What tool will be used for making decisions, executing tasks and managing workflows?
How to effectively use automation
Automation can be a powerful tool, but IT leaders should be careful not to automate just for automation’s sake or needlessly push automated tasked-based processes into end-to-end automation.
Automation isn’t always a sound investment, especially if the process requires infrequent work and minimal time. In a previous role, I helped roll out a new accounting system, and as part of the new technology, the company’s CFO wanted to automate a complex financial report. My team mapped out how to orchestrate the automation of this report, estimating weeks of IT work and a significant financial investment. Luckily, before getting to work, we asked the CFO how often the report had to be run. Just once per year. That ROI didn’t make sense, and we used our resources elsewhere.
Sometimes simple, task-based automations are sufficient, and businesses don’t require more advanced solutions. I have gone into companies where they can have 3,000 scripts running, doing all kinds of things: pinging a router to make sure it's up and running, querying a device to check on its memory or checking a server’s load. And if these automated tasks don’t need to talk, then task-based automation could be enough.
For many companies, end-to-end automation can offer significant value with very little effort. At my company Windward, we survey customers after every project, but our manual process was allowing 70 percent of these surveys to fall through the cracks. So we orchestrated an end-to-end automated survey process. When a project in our CRM system hits 85 percent completion, our automated system creates a questionnaire, texts a link to the customer and a Slack integration notifies the project manager of the activity. When the customer completes the survey, the CRM system receives the data, and depending on the score, the customer gets one of three email templates from me, thanking them for their time and inviting them to catch up. The automated system then emails my assistant who schedules a time to meet with the customer. And here’s the value: we surprise and delight our customers by never missing a survey and enhancing the human-to-human relationships with our customers.
As the vast world of automation matures, more businesses are adopting automated processes in some form. But before embracing automation, stakeholders need to ensure there is a solid business case behind the technology and then follow practical steps for creating value-driving end-to-end automated processes. Then, businesses can start adding time back into their employees’ days, providing continuous customer service, streamlining processes, saving money, enabling company growth -- the possibilities are endless.
Sean McDermott is founder and CEO of Windward Consulting and RedMonocle. He also acts as lead researcher at Helix Market Research. Sean previously acted as founder and CEO of RealOps Inc., the pioneer in enterprise management run book automation solutions, which was acquired by BMC. Sean’s curiosity for advancing technology began at his first job as a network engineer/architect installing and managing the first private internet for the U.S. Department of Justice. At a time when the internet was just taking off, Sean was at the forefront and has continued to be on the cutting edge of technology with the development of Windward and RedMonocle. He is an advocate for business leadership strategies and shares how other entrepreneurs can align passion and action on his blog, Wheels up World.