Hello, Microsoft, A "Guided Process Tour" is not Process Discovery.
Microsoft recently announced its Power Automate and robotic process automation (RPA) tools to define and automate processes.
It's great to see that Microsoft is focusing on BPM. We welcome Microsoft to the arena, but here is what they missed from what makes a robust process discovery tool.
In this post, we will discuss what it takes to discover and optimize business processes effectively, and we'll point out how Skan.ai approaches the job.
The Key Missing Ingredient from Microsoft's Automation Suite
First, the company announced its process advisor. It's the feature of Power Automate that helps one discover the steps that comprise a business process, which might be suitable for automation. It also added RPA enhancements to Power Automate. RPA uses bots to eliminate repetitive tasks and automate specific steps in any given process. Their release focuses on RPA features that help find the repetitive tasks most suited for automation.
While Microsoft's no-code approach to automation will appeal to some companies, there's a great deal of work yet to do in discovering, defining, analyzing, and optimizing business processes that Power Automate does not address.
Microsoft's approach to understanding the process can be termed a "Guided Process Tour" - as in a participant navigating the process's path as a "show and tell." It is similar to what a tour guide does – facilitate a tour of what they think are the highlights and lowlights of an attraction. Alas, this does not work well based on Polanyi's paradox that people know more than what they can tell. Furthermore, when something is being "recorded," people work how it is in the standard operating procedure and not necessarily as they do when no one is watching.
Process discovery is the order of magnitude different and complex than a "guided process tour." Deciphering and depicting a process across multiple team members, terminals, time zones, applications, and workflows is like finding patterns from the random grains of sand on a beach.
How Skan.ai Discovers Processes?
Skan.ai has focused its efforts on visualizing human work and building a metamodel of that work. Unlike Microsoft's Power Automate, we don't assume that workers can accurately describe the flow of activity that comprises a given process. Neither do we urge you to hire a battalion of consultants to interview your workers so they can come up with process flows.
For example, asking an accounts payable worker to describe the process for paying an invoice might reveal four necessary steps: (1) receiving the invoice, (2) filing it in the Accounts Payable system, (3) getting approval to pay, and (4) paying the invoice.
However, asking people to define such a business process can reveal only part of it. There are almost always other steps required that are easy to overlook — such as checking for prompt-pay discounts, confirming the goods or services delivered are acceptable, etc. Likewise, asking more than one person in accounts payable to process invoices can result in multiple answers and process flows.
Skan monitors each user's screen to determine what users do as they process an invoice. Whether they use Macs, PCs, Citrix, or green-screen terminals on an AS/400, Skan.ai software captures what happens on the screen and the step-by-step process in its entirety. That process can then become the approved method for processing invoices once management makes any necessary enhancements and improvements.
With a given process fully documented, Skan.ai also watches what each worker processing invoices do so management can spot "process drift," which occurs when users, intentionally or otherwise, insert new, often inefficient steps into the process. This continuous monitoring of dozens or even hundreds of cases gives supervisors the intelligence they need to eliminate inefficiencies, and we call this "operational intelligence."
In contrast, Power Automate focuses on finding steps in a process to automate, then provides a way to automate those that qualify. I'd liken that approach to a surgeon wielding a scalpel, looking for something to cut into and repair. While that's useful, it cannot build the metamodel of human work that's the foundation needed for genuinely effective business process management.
Even beyond business process management, knowing the metamodel for each process is foundational to any digital transformation effort. Understanding how humans get work done reveals where digital solutions can increase efficiency, reduce costs, and deliver a more positive customer experience.
And Then, Finally…
Then, once processes are thoroughly and accurately defined, one can look at how to improve them. However, automating specific steps is just one way to improve the process. Some improvements might require a change to the underlying "tooling" and some may require ‘outsourcing” and yet others may need to be “eliminated” or “re-engineered. Or perhaps human bottlenecks impair specific operations which is more of a capacity calibration issue.
The Discovery of the actual steps in a process opens possibilities for process improvement beyond just automating certain steps. Automation is essential, but it is but one intervention. A complete metamodel of work and the resulting process intelligence carries much more weight than Mickey Mouse automation. This is something Power Automate cannot do.
No matter where you are in a digital transformation effort or how you're attempting to optimize business processes, process discovery is a foundational step. And Skan is the epitome of process mining's next evolution, incorporating computer vision and deep learning to digitize known, unknown, visible, and invisible processes into a process abstraction. The resulting operational intelligence allows for modeling, simulation, predictions, and conformance to power the digital enterprise.
So, we want to thank Microsoft for bringing the spotlight to the category, but humbly submit that intelligent buyers will see the gulf between a “guided process tour” and “process discovery.
Please Contact us to learn more or to schedule a demo