Traditional process mapping techniques fail to account for the "invisible enterprise," which refers to the deviations that arise organically within real-world business processes. These deviations can lead to suboptimal automation efforts and unsuccessful transformation initiatives. However, by leveraging artificial intelligence-powered process mining, organizations can capture these variations, thereby enabling the creation of more accurate process maps.
Are your automation efforts and transformation endeavors proving to be futile and benefits seem elusive? Your company may be the victim of ignoring the “Invisible Enterprise.” Please read on to explore the concept and how you can avoid the pitfalls that come with not realizing the reality of the invisible enterprise.
We all agree that business processes are the core of any enterprise. A logical and hierarchical grouping of Value Streams, Business Processes, Workflows, Activities, and Tasks constitutes a complex web of interdependent, at times invisible, and often intricate patterns that are indeed akin to the central nervous system.
Process charts or maps are nearly a hundred years old concept. Lillian and Frank Gilbreth introduced the concept in 1921 in a paper titled: “Process Charts: First steps in finding the one best way to do work”
And how do we try to map these complicated process flows? From the days of Taylor/Gilbreth and the infamous Time and Motion Studies to the process re-engineering revolution, the concept of Kaizen, and to the currently in vogue RPA (Robotic Process Automation), the modus operandi has been unleashing an army of consultants to map the processes.
This is how it begins with any transformation endeavor – whether it is functional transformation such as finance (Record to Report) or HR (Hire to Retire) or any number of digital transformation endeavors to reinvent a range of functions. A team of analysts/consultants conduct a current state assessment, then envision a future state, and craft a transformation roadmap.
An integral part of the current state assessment and also the future state target operating model design is the mapping of the relevant processes.
The typical enterprise process mapping exercise involves one or more of the following:
Then the team of (human) analysts’ whiteboards the process step and then constructs a process map – for example, a swim lane chart – and possibly other allied artifacts such as a dataflow diagram, a state-machine diagram, and the like.
Most of the process maps tend to be of two varieties with one fatal flaw pervading most of these endeavors.
One set of process maps is too high-level with a few boxes and arrows and does not capture the vivid details and underlying flows well. The other set of process maps tends to be too much in the weeds with a lot of noise missing the signal. The fatal flaw – that afflicts both varieties is that almost all of them focus on the happy path plus perhaps one or two exception paths. This flaw is one of the root causes of suboptimal automation and failed transformations.
In essence, the human-generated process maps tend to miss the “invisible enterprise.”
The invisible enterprise is an amalgamation of the drift and nuances that result in minor yet significant divergence to the enterprise business processes that over time take a life of their own and result in hundreds of variants that are typically ignored by traditional human-led process mapping.
Like “Shadow IT,” the Invisible Enterprise is a practical reality and executives in-charge of re-engineering, automation, and transformation efforts can ignore at their peril.
The Invisible Enterprise is not insidious or inherently evil, but it can turn into one. The process variations are a result of suboptimal technology enablement, new regulations, and standardized processes, not keeping pace with business requirements. As employees get to resolve issues that are outside the standard straight-through processing, these process variants tend to germinate and over time perpetuate.
Necessity is the mother of Invention: Employees don’t wake up one beautiful morning and decide to change up the process. Instead, most process variants are a result of difficulties of the situation or business needs that require a slight twist in the standardized procedure.
Regulation-Mandated Process Variations: The regulatory burden is ever-increasing, and companies often tend to put a band-aid on a system and create a patchwork process to meet the mandate. Regulatory compliance often creates process variations and exception flows.
Inflexible and Rigid Process Flow Configurations: When companies purchase (or build) software systems, the configurations, and customizations often focus on the happy path, ignoring the edge cases. As the volume and value of the edge cases grow, they tend to create process variants.
Suboptimal Automation: Without an accurate picture of the underlying process and not accounting for process variations causes the automation efforts to take considerable time and rework. It is not unusual for the hardening of the automation to take somewhere between six and nine months. Without a transparent process model, bots may run wild performing tasks, but true process automation will remain elusive.
Piecemeal Transformation: Today, many companies are finding their digital transformation efforts are failing or falling short of expectations. Lack of process understanding, a tool-first approach, and lack of strategy are perhaps the root causes of this. Without a complete picture of the process, any re-engineering and re-platforming efforts, which are fundamental to a transformation will end being a patchwork and a paint job.
So, how do we overcome the challenges of traditional process mapping and include the invisible enterprise for accurate and comprehensive process maps? Glad you asked. An AI-powered method to capture the Invisible Enterprise will help enterprise build a factual foundation. Synthesizing computer vision, machine learning, and advanced data science, process mining and process discovery digitize every machine and human interaction and from the digital traces of human work compile a process metamodel.
Intrigued? Interested? Drop us a line, and we’ll be happy to help.