Digital transformation can be difficult for businesses because it's hard to understand exactly how work gets done. Process mining uses data from computer systems to analyze workflows, but it can miss human involvement. Skan offers process discovery, which uses AI to observe user interactions and create a more complete picture of business processes. This can help companies identify areas for improvement and make better decisions about digital transformation initiatives.
Going by years of historical data and case studies, it’s a well-known fact that transforming a company in fundamental ways has always been a big challenge. Digital transformation is no different, especially in the context of new-age complex organizations. This article looks at the interplay between the complexities of digitally transforming a company, and the critical role process mining and process discovery play. Let’s begin with a short recap of the digital transformational efforts so many companies are committed to today.
The pace of technological advancement in recent years has opened businesses to entirely new tools that promise to increase efficiency and productivity, to reduce expenses, and to enhance customer service and satisfaction. Artificial intelligence, machine learning, computer vision, natural language processing, chatbots, IoT, cloud services, 5G, mobile services, and more have given businesses new capabilities that, it’s hoped, will accomplish all of those goals.
As companies strive to keep up with ever-growing customer expectations and addressing other business challenges, they are bringing a constant flow of new technology and changes in the business environment, thereby finding themselves at the epicenter of what’s come to be known as digital transformation. The Australian telecom firm Telestra, for instance, surveyed more than 3,800 global executives and decision-makers in 2019 to reveal what works and what doesn’t work as companies move toward digital transformation. Telestra found only one in five companies (21 percent) consider themselves “digitally mature,” while just 24 percent amongst them being able to claim that they could integrate a transformation throughout the organization.
Management consultants at McKinsey further point out that only about one in six companies (16 percent) report that their transformational efforts improved business performance and that such improvements were expected to continue over the long haul. Another 7 percent said that improvements were obtained in the short term but were not maintained.
It’s rather strange that this is when numerous multinational consulting firms, think tanks, pundits, and scholars have published reams of advice on how to succeed at digital transformation. Much of that counsel focuses on a few key areas that include technology, along with the need for C-suite support and engaging stakeholders throughout the organization, establishing clearly defined goals, and acknowledging that transformation is not a “once and done” project. At the same time, even with these clearly defined goals, there is a definite gap that’s leading to rather limited success with the overall transformation in organizations. We at Skan believe that this gap is what we need to bridge, and adding Process Mining and Discovery to the toolkit is the key here.
Wil van der Aalst, a Dutch professor, pioneered process mining. In recent years, process mining has taken hold throughout Europe and is beginning to draw the attention of North American businesses as well.
Before we go further, just to elaborate, process mining is an analytical method that extracts information from event logs (and other sources we’ll discuss below) intending to discover how processes work, as opposed to how one thinks they work. Then it enhances those processes basis what the analysis reveals; and finally, monitors how each process flows over time, with a particular eye toward deviations that could affect the desired business outcomes.
For instance, a company’s accounts payable department might notice that it frequently misses out on prompt-pay discounts. Analyzing the step-by-step processes involved in securing those discounts, one can discover where, how, and why they’re being lost. This understanding can give management the knowledge it needs to improve that specific accounts payable process using facts rather than conjecture, speculation or guesswork.
For example, a major pharmaceutical company decided to grow its business by undertaking a digital transformation journey. The company was already heavily invested in several enterprise-level software suites — SAP’s ERP system, Salesforce CRM, and others. Those software apps supported hundreds of business processes ranging from finance to manufacturing to sales and marketing. Each of those generated log files that documented virtually every step taken in each business process.
Once the company decided what area of the business would be first on the transformation calendar, it considered holding workshops that would bring relevant people into the room to discuss how the current process was working, where it might be improved, and how best to modify the process to achieve more desirable business outcomes. However, the proposed workshop leader pointed out that it wouldn’t be possible to bring everyone into the workshop who might have relevant input. Many of those people were scattered around the globe so that even a live webinar wouldn’t work due to time zone differences.
Their ah-ha moment came when they began to study the log files that had been continually documenting details on how business processes are run.
What, they asked, if we could look into those files, then use them as the basis for understanding our existing business processes? We wouldn’t need to bring crowds of people into a workshop where each individual would voice their own perspective, including their unique ideas, biases, and conjectures. Instead, we could make a graphic representation, perhaps a flowchart, that illustrates how the process works.
By analyzing or “mining,” logfile data, you can avoid the subjective opinions a workgroup, a workshop or an outside consultant delivers, replacing it instead with hard data that’s been accumulating every second of the day for weeks, months or years. Moreover, process mining provides a lens that you can focus on any specific business process you wish. For example, you can look at those involved between the time you order from a vendor until that vendor is paid. You can just as easily mine those used for the steps involved in the order-to-cash cycle, or virtually any other business process.
Yet, there’s more. Process mining based on log files shows you what has transpired in the digital world; it shows you what computers have done in support of your business processes. However, have we paid heed to the people who take part in the business process? How do we factor in what Alice or Bob have done and how they’ve contributed to a business process? Log-based process mining provides the visibility only between two committed states of the process but misses entirely out whatever happened in between those two committed states. Process mining entirely depends on the granularity and the interval of the logs which underlying system is writing.
The answer lies in “task mining” or “process discovery,” which brings the human element into focus by adding user interaction data. This addition to process mining uses computer vision to observe the human-machine interactions that occur as a process is being executed.
For instance, if Bob in Purchasing needs to order widgets, he’ll use software that leads him through the steps of creating a purchase order. Computer vision captures those steps as Bob enters the vendor name, the items to be ordered, their prices, the company’s payment terms, the ship-to address, and other details. Those user interactions add context to the data pulled from log files. Then, with the help of AI routines, the human interactions and log file digital actions are merged into a far more robust illustration of what happens when Bob prepares a purchase order.
In this example of task mining, you might learn that it takes Bob four minutes to check the prices he paid on his last widget order and to consult the contract terms between Bob’s company and the vendor. Only after he’s confirmed those can he issue a new P.O. Task mining, in this case, points to a four-minute inefficiency that can be addressed by adjusting the process itself.
Overall task mining or process discovery helps uncover the tribal knowledge or cheat codes being used by the process users. It helps highlight the critical aspect of human interaction with digital systems.
Skan Process discovery equals computer vision + machine intelligence. This is the approach we take at Skan. We capture the user interaction data using computer vision to augment the traditional log file approach. The result: an easy and intuitive way to understand your business processes and to better prepare for digital transformation initiatives.
Our Skan process mining software — Skan CPX — helps you with process discovery and understanding where to focus your digital transformation efforts. It’s an AI-powered platform for uncovering, untangling, and unleashing enterprise business processes. Skan CPX delivers immediate and enduring results for enterprises with complex processes and expansive business and IT landscapes.
The Skan CPX artificial intelligence component, the Virtual Process Analyst, observes and captures every human and machine interaction unobtrusively and learns continuously. And it does so without systems integration or the need for APIs. Skan CPX uses computer vision and machine intelligence to power process capture.
No matter where you are in your transformation journey or how large your company, we can give you a clear and unambiguous view of your existing processes so you can succeed in your digital transformation. We call it Google Maps for the enterprise digital transformation journey. You know where you want to go, but we help you understand where you are at the moment and what’s the fastest and most efficient way to reach the destination.