Robotic Process Automation is revolutionizing industry by handling repetitive back-office operations once done only by humans. Should you join the early adopters?

Automating processes through the use of software has gained significant traction in recent years. It offers companies the ability to dramatically reduce cost, eliminate errors and provide “instant” responses, and all of this without making considerable investments in new IT systems. Critically though, only when a process is stable and predictable is it actually “fit and ready” for automation.

If the process is unstable a lot of time and money will be wasted fixing issues that could and should be solved before attempting automation. Here’s how to stabilize your processes.

What is Process Automation?

Robotic Process Automation (RPA) involves the use of software (robots) to automate steps in a process previously executed by humans. Robotics has the potential to revolutionize the way service companies complete their administrative and back office operations. Robots can radically lower cost, shorten cycle times, achieve regulatory compliance and deliver consistent quality to customers. Some early adopters are already benefitting from robots, while others are finding they are not the anticipated silver bullet. We know clients who have postponed automation initiatives after discovering their processes were just not fit and ready.

The appeal of RPA is clear: robots do not get sick, take holidays, get bored, lose concentration or leave for other jobs. Furthermore, a standard work pattern does not limit them, so they can be used when required. Robots are not, however, perfect; they can do what they are told to do repeatedly and well, but they have to be told precisely what to do for each and every circumstance. They lack common sense, so for example, intuitive judgments made by people have to be understood in detail and codified so they can be taught to the robot. This includes processes (including each and every alternate path), procedures, and decision criteria.

Are Your Processes Stable?

In reality, many companies struggle first with standardizing their processes (and their inputs), and then making them stable and predictable is even harder. Organizations that have properly designed and clearly defined processes often still see differences in the way people execute process activities, sometimes due to personal preferences, but often due to variations inherent in the overall system.

While people can intuitively adapt and deal with variability in the system, robots do not adapt unless programmed to do so, and with each additional variable the level of complexity increases. If the robot is not programmed to handle all possible variations, the result is a lot of exceptions that need to be passed to a human to complete. In the words of the great American engineer and businessman W. Edwards Deming, “If I had to reduce my message to just a few words, I’d say it all had to do with reducing variation.” In this context, process stability is a prerequisite to automation, not a result of it.

How Do I Stabilize My Processes?

Here are some practical steps organizations can take to stabilize their processes:

1. Process Design and Definition – Ensure processes and inputs are designed, defined and documented from the customers’ perspective. Include what criteria will be used to measure performance.

2. Process Conformance – Ensure all team members understand, are competent in, and consistently apply the defined process.

3. Process Stability – Identify and remove all special causes of variation, e.g., one-off random events that destabilize performance and create variability. Once stable, process performance becomes predictable.

4. Process Improvement – Identify and implement continuous incremental improvements to the process. Take steps to ensure improvements are incorporated into the standards and the higher level of performance is consistently maintained.

5. Process Re-engineering – Redesign and redefine the entire process to make a step change in customer experience and operational performance. This is the point at which the introduction of automation is relevant.

The re-engineering may include the application of “ECRS” thinking:

1. Eliminate – Which processes, activities or steps can be removed completely?

2. Combine – Are there any processes, activities or steps which should be aggregated?

3. Rearrange – Should the order of processes, activities or steps be resequenced?

4. Simplify – Can the remaining processes, activities or steps be simplified?

Automation Complements Improvement

Automation has many obvious benefits, but it should not be seen as a replacement for excellent design, implementation, measurement, management and continuous improvement of processes. Some organizations are seeing automation as a way out of poor process design, implementation, measurement and improvement. Automation may temporarily mask many deficiencies, but ultimately an inferior process is always going to be an inferior process, and in this context automation is not the answer. As Bill Gates noted “the first rule of any technology is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient process will magnify the inefficiency.”