Benefits of Aligning Robotic Process Automation and Continuous Improvement
Published on September 5, 2019 by Mark Abrams LSS Black Belt, Managing Director, Polaris Process Innovations
Robotic Process Automation is an effective tool for helping businesses improve their operations by leveraging “Robots” or technology to automate mundane and manual tasks that are performed by humans. This allows a company to reduce the time needed to process transactions and reduce operational costs. In general, there have been noticeable improvements in both operational cost and cycle time reduction as well as improvements in overall operational quality that companies have seen.
After watching how RPA technology and operational successes have progressed over the last two years, I think that we have seen solid success in automating simple processes. The next phase that we are now entering into with RPA is implementing automation where complex decision making activities and analysis are required within the process. This phase will present significant challenges for everyone including Developers, Business Analysts, Process Owners and Stakeholders in delivering true breakthrough improvement for businesses implementing RPA.
Identifying Process Complexity
Process complexity can present itself in a variety of ways, making it difficult to identify. There is usually no way to tell where a process should be segmented or automated when first documenting steps in it. This is problematic because if the complexity isn’t identified and addressed early on in the development phase or before, the opportunity for costly delays increases exponentially.
Process complexity can be encountered in multiple ways:
A seemingly large amount of process steps that an individual must take to complete an individual process.
Multiple inputs are required for a person to complete their analysis of a transaction.
Two or more processes being managed and measured as one.
There are other warning flags that Process Owners will give their interviewers, and these too are vague and innocent sounding as well:
We need more people but we haven’t been able to get the req’s approved.
My people are burning out or are stressed out from their workload.
We’re running a lot of overtime.
We need better technology to perform our job.
In these scenarios, the process owner will most likely not realize what is really wrong with their process. To the untrained ear, these symptoms immediately scream “automation is needed here, right now!” which is most likely the reason why the process was selected for automation in the first place. However, to someone attuned to identifying these subtle symptoms, these are indicators of other deeper, operational problems that have less to do with the need for better technology than people might think.
Understanding the process:
In the rush and excitement associated with starting a new automation project, it’s easy to jump right in and start interviewing people and documenting everything in sight. Forgetting for a moment that we need to document the business case, use cases, requirements and the steps in the process, let’s step back and focus on several key areas, namely metrics, process documentation and transactional data generated by the process. These three areas will provide us telling information about the process.
From my perspective, when reviewing a process for the first time, the focus should not be on the developer or any technical system (application). My preference is to document the process at a detailed level so that I understand it and can explain it back to anyone. This is different from the “technology first” approach that has us examining existing technology as the problem and applying new technology as the solution. To me, it’s always about the flow of the transaction and NOT the technology. In order to help frame what this means, follow the basic wastes of Lean Manufacturing. Simply observe the process, document the steps and pain points that are encountered in each step. This will be helpful for you AND the developers later on and besides, people love to tell you about the problems that they have in their process!
Wastes in the process
While it is helpful to stay agnostic of process improvement methodologies as one isn’t necessarily better than the other, Lean Manufacturing does provide very helpful ways to categorize various problems within the process. Granted, these wastes do stem directly from manufacturing but still apply to transactional and service operations. I’ve bolded the categories that I see most frequently in back office processes.
Transportation (moving something unnecessarily)
Inventory (transactions or materials piling up)
Motion (making unnecessary movements between different applications, etc.)
Waiting (waiting for information to be supplied to complete a transaction)
Over Processing (doing more than required for a transaction)
Over Production (unnecessary reports or documents filled out)
Defects (doing something that causes a transaction to be returned by another process or customer)
Skills (wasting people’s talents / skills by having them perform useless or wasteful tasks and not utilizing them to improve their work)
The reason for using waste categories alongside pain points is that by focusing entirely on pain points, we miss the underlying operational problems we are trying to solve. For instance, “inventory” building up in a queue does not necessarily point to technology as the cause or the solution. Other operational problems such as scheduling staff incorrectly and new products introduced that were incompletely designed and or rolled out are two very common issues encountered. If schedules can be easily adjusted and reduce queue volume and ultimately the time it takes for a transaction to be completed without needing to spend money on automation, then that a positive and immediate return on investment.
Real Life...
Let’s examine two real life scenarios where technology and process improvement came together to effectively solve operational problems.
In the graphic shown above, we were asked to improve the sales process for a technology company. The “predetermined” solution was automation because that will undoubtedly make the process go faster. A “spaghetti map” is an effective way to show motion in the process as it is extremely hard to reflect an agonizingly slow process on a process map. The lines represent email exchanges for a particular client, with the red line (approval) being the most important. The black lines are people who were cc’d on the email threads so “they were aware of what was happening.” The red and black lines represent the first part of a transaction that took more than 7 days to finalize while taking less than 2 hours of actual time to complete. The blue lines represent an additional 4 days for people to figure out how to charge for a particular piece of functionality. Remember— there is a customer attached to this transaction!
Reflecting on the chart above, the thought that comes to mind is that this is what causes our inboxes to overflow with email! If one asks: “Why does it take so long to get an answer to a particular question?” The first thought that comes to mind is not outdated or under performing technology, but what information are they missing to complete the sale? The potential root causes will most likely group together in process and change management areas rather than technology not working correctly.
In the chart above, a process was identified as a candidate for RPA was experiencing long lead times, excessive overtime and other operational challenges. Before automating the process, analysis was done on the operational data which was able to describe the pain people in the process were feeling.
This chart shows the impact of volume and cycle time. The blue line represents the arrival rate of transactions in a process (inventory.) The orange line represents the overall time it took for the people in the process to complete the transactions (waiting.) The variation in the blue line is normal (surprisingly enough) while the cycle time (orange line) shows a gradual build and several large spikes and valleys. The peaks and valleys translate to the team “cleaning out their queues.” Further analysis (not shown) revealed other causes for the long lead times, specifically pointing to how the operation was scheduling labor.
In my experience improving processes over the last 16 years in Back Office and IT environments, business stakeholders mistakenly jump to automation as the solution in an effort to improve their company’s ability to deliver quickly, efficiently and cost effectively. Yes, there are times when technology as the solution is the right answer. A lot of other times however, it’s improving the process that is the solution. Ultimately, understanding the process more thoroughly is what is needed to fully realize the benefits of automation.
Join me for part 2 of this article shortly. If you have any questions or comments, please feel free to reach out to me and I would be happy to answer them.
Mark Abrams, LSS Black Belt, Managing Director, Polaris Process Innovations
Mark has been improving processes for the last 15 years in a variety of areas including Hotels, Manufacturing, Laboratories, Supply Chain, Procurement, Professional Services, Cyber Security, HR/Payroll and Fraud Prevention operations.