By Jason Wickman, vice president of operations, Process Analytics Factory (PAF)
Process Analytics Factory's Jason Wickman
Robotic process automation (RPA) is quickly becoming the rising star of the business-optimization world. The challenges brought by the coronavirus pandemic have only accelerated the demand for workflow automation in industries of all kinds—and with good reason. When implemented properly, RPA can be a key technology for optimizing efficient, error-free and transparent business processes.
But despite its many benefits, approximately 30% of RPA projects fail because the automation tools and/or processes are not well understood. RPA works best when applied to increasing the efficiency of repetitive processes. It can’t fix a broken process or “learn” how to react to unexpected events. According to Forbes, “RPA’s ‘robotic’ nature means that it is geared toward well-defined data formats, steps and outcomes. Throw in unstructured data or process variation, and RPA struggles at best or breaks at worst.”
That’s where process mining can help. Process mining provides insight into how processes actually transpire, how they deviate from the ideal model. It informs decisions as to which problems occur and which optimization measures should be taken.
For example, most companies have well-documented procure-to-pay business processes. In reality, however, anomalies arise during implementation. Process mining harnesses the business’ own data to identify, for example, whether releases are made in the specified timeframe, how often the approval process is adhered to or bypassed, how many variants of the release process exist and whether delays are adversely impacting the business process.
When used in combination with business-analytics tools, process mining provides visibility into the totality of all orders for a defined period of time, right down to the examination of individual orders. Unlike task mining, which looks at an individual’s interactions with business processes, process mining can gather valuable insights across platforms, applications and even departments.
Leveraging data for successful RPA implementation
Today, it’s estimated that only 23% of users have access to analytical insights in their business applications. Worse, Gartner predicts that only 20% of analytic insights will deliver business outcomes, noting that “for many organizations, activation, or the art of leveraging data to do something meaningfully different in the market, is the missing piece that bridges the divide between insight and business value.”
The key to successful RPA implementation is a solid knowledge of the processes being automated. Six in ten leaders working on RPA projects told CIO Dive that “a strong understanding of the processes being automated helped their projects thrive.” However, in our experience, the cost of identifying, analyzing and documenting processes can often account for well over half of the total effort of automation projects.
A clear case for process mining
Process mining gives businesses the transparency and clarity they need to identify those processes that best lend themselves to automation. Interviews and random sampling are not sufficient to identify all possible variants in a process chain. With process mining, organizations eliminate the “hit or miss” approach, which in turn can reduce RPA costs.
RPA adoption is across industries and in businesses of all sizes, even in today’s market. Organizations need transparency and clarity about a process before they can successfully automate it. Otherwise, they miss out on variants, compliance issues and other exceptions and invest in a project that, at best, will not bring the expected results, or, at worst, is doomed to fail entirely.
Process mining is a valuable asset to any optimization strategy. It can help unlock the potential of automation and make sure your organization is getting the most out of its RPA measures.
