Welcome to the “great automation debate” centered around one pivotal question: should you automate a bad or broken process? Almost universally, the answer is no. Many process experts even claim it would be ridiculous. Why would you want to automate something that is broken?
With great respect for my process expert colleagues – please hold the phone! I’ve looked long and hard at the benefits and tradeoffs of automation, and I suggest we take a different stance. I stand with some of our most determined, persistent and successful clients when I say that the most prudent answer to the question about automating a bad process is an ardent, “Yes!”
Let’s face it. Bad processes impact humans more than good processes. You never hear someone chatting at the water cooler about a good process. It’s always the problematic ones we hear about. Bad processes create inefficiency and require extra effort to handle work-around procedures, rework cycles and duplicate tasks. They also increase the time and energy spent escalating issues to more expensive humans. When a bad process impacts many humans, the costs and negative consequences grow exponentially.
So Why Automate?
The answer is simple – to drive dramatic savings and benefits faster than traditional re-engineering will. The following are a sampling of the benefits.
- “The flashlight effect.” At the dawn of the digital labor age, I worked with one of the largest insurance providers on the globe. It had an innovation budget set aside to explore automation even though the benefits of automation at that time were still soft and unknown. The benefit of the automation exercise was to “shine a light” on the company’s ineffective processes and excessive costs. The probability of fixing a bad process is much higher when you shine a light on it by automating it.
- Measurement. How often does someone initiate a work-around for a bad process in the enterprise? Imagine an HR onboarding process that incorrectly orders business cards for a new employee. A seasoned HR rep understands that the process of ordering business cards does not properly identify the employee’s correct business line. (S)he initiates a workaround by calling the new employee ahead of time. Then the employee acts on the new information provided by HR and manually enters his/her correct business line but inadvertently makes a spelling mistake while entering the information manually. The business card vendor prints the business card and charges the employer, but the new employee has incorrect business cards. Now consider this example when there are 30 HR reps and 10 use different workarounds for this broken process while 20 let the process break normally. How many failures happen per day? No one knows! But, as soon as you automate the process, you immediately begin to gather facts that help you determine where it is broken. Very quickly, this provides data that can be tied to value that, until the process was automated, was completely unknown in scale or impact.
- Speed to fix. The process of automating a bad process consolidates all the elements of the process in a single place. This has two important consequences. First, it means that remediating a bad process, even incrementally, is much easier once it’s automated than trying to impact the bad process in its native, analog state. Second, today’s robotic process automation (RPA) tools make it easier to automate and improve the process than to re-engineer, remap, document and update the process and then generate the necessary screens and reports to train employees.
- Potential for improvement. Automating a process and deploying a robot makes it much easier to improve the process as you go. Most processes in large enterprises touch more than one technology and more than one stakeholder. If one RPA bot can act as the process executioner and governor, then successive improvements in the automated process will reduce complexity by orders of magnitude. In the HR example above, the HR rep would never have to initiate the workaround and the “robot” could value-check the business cards against known company standards before issuing the order to the vendor – reducing human intervention and improving the quality and consistency of process execution.
What if you don’t automate the bad process? This is the trap that is easiest for enterprises to fall into because not automating means retaining the status quo. It “feels” like nothing happens, but something is happening. The business cards still get ordered twice, the business gets billed twice and HR adds to its reputation for being a large cost center without visibility to explain why. Moreover, the C suite has no idea there’s even an issue that could be addressed.
Bring those bad processes to the bots! Let ISG help you identify processes that are ripe for automation and build a business case to get you started. Contact us to find out how.
About the author
Darryl Hamer is the Business Development Executive for ISG’s Automation practice in North America. His role is to assist prospective clients to understand how automation drives value and benefits and to plan ISG Automation Advisory Services to support a client’s full automation needs from Robotic Process Automation to Cognitive/Artificial Intelligence to successfully completing digital journey.