Process Mining a Quintessential Tool For Next Level Initiatives

The typical approach to success in business, is to plan, execute as per plan, monitor the achievement of the plan, and take corrective action where necessary. To operationalize this, most organizations create a dashboard of key performance indicators (KPIs) to keep an eye on how the organization is performing. KPIs measure the business outcomes.

For example, the OTIF (On time in full) ratio indicates the extent to which an organization has been fulfilling their orders on time and in full. A lower OTIF is bad and a higher OTIF is good. But viewing the OTIF rate as an independent KPI has its own problems. An improved OTIF could have been achieved by building up undue inventory. This then is improvement of one KPI at the cost of another. On the other hand, a poor OTIF rate may be on account of less accurate demand forecast.

Hence it is important to analyze KPIs not as independent yardsticks of performance, but collectively along with associated KPIs to have the complete picture.

But when we look at KPIs, they represent mere symptoms. They only provide an indication of how and what is happening. They only provide a partial view of the whole picture. They do not reflect the underlying causes. And if the causes cannot be understood, it is difficult to fix the problems. The management cannot take concrete action to resolve the issues wholesomely.

Most KPIs are impacted by the way the underlying process is carried out. The efficiency of the internal processes has a great influence on the performance of the KPIs. In the same example, it could be that the time taken to release credit blocks is the biggest stumbling block for delivery. Or it could be an approval for the price that is always delayed. Unearthing such root causes help in effectively managing the performance of the KPI.

These Process Efficiency indicators (PEIs) are thus the diagnosis of the problems in many KPIs. This in turn means that looking at KPIs along with PEIs is a more holistic approach to solving key business problems. However, these PEIs are hidden in the actual processes and need to be discovered.

Process mining is the art and science of unearthing such gems from the underlying processes. It involves extracting the readily available knowledge from the event logs of information systems and using that data to discover, monitor and improve actual processes. The most obvious imperative for process mining is to measure and act on PEIs. However, process mining has become a tool to look at topics much beyond business process improvement.

As a first step, process mining helps in understanding the actual processes being carried out as against the prescribed way. It has been proven that the actual variations in process flows can not only surprise you, but also provide a deep understanding of the actual needs of the business. Not all variations are necessarily deviations.

Secondly, with process mining, we can also grasp the bottlenecks in the process – which steps take the most time, what are the causes of long throughput time, how many times a step gets repeated in how many cases, are there anomalies in the sequence of steps and if so in how many cases and so on. These analyses help build up a strong understanding of the reality of how the processes work rather than relying on the presumed or prescribed definition.

Thirdly, when these above-mentioned insights are analyzed, we can identify execution gaps in the processes. For example, in the Procure to Pay process, we could discover that in a significant number of cases, there is an additional step of “Price change” after the communication of the Purchase Order to the Vendor. This could indicate that the price for a particular combination of materials and vendors is not clearly fixed and that results in an overall delay in supply. A remedial measure could be to identify such cases and work on entering into a price contract with the vendor(s) for those materials. Hence simple business measures to fix the gaps could significantly improve the process efficiency and result in a better KPI elsewhere.

Lastly, a deep analysis of the processes with a digital lens would also throw up opportunities for digitalization. The customer master creation may be a manual process wherein details provided by the sales team are checked and uploaded by the finance team into the ERP. This is a classic opportunity for complete automation using Robotic Process automation. In the Accounts Receivable process, there could be frequent delayed collections, indicating the need for an AI/ML solution that identifies and proposes specific invoices for which a collection reminder could speed up the payment. Such an analysis of the processes thus helps in defining the digital transformation roadmap for the organization.

Process mining solution providers such as Celonis (the current Market leader in this topic) provide complete Execution Management solutions.

In summary, the case for Process mining is directed towards the following objectives:

1. Business Process improvement

2. Conformance and controls (for internal and external audit)

3. Execution Management (resolving execution gaps)

4. Identifying Digitalization opportunities (Automation, AI/ML, Chatbots, etc.)

Process mining, thus, is an important means to progress specific functions as well as the whole organization to the next level of functioning. In this blog, I spoke about “Digitalizing the Finance Function to Thrive in the Next Decade”. Many of the levers mentioned in that write-up can be facilitated by process mining. The same can be extended to Supply Chain, Sales, Procurement, and other important functions of the organization.

With computerization in general and ERP in particular, having become ubiquitous and more of a hygiene factor, it is imperative that organizations shift gear into the next stage of Digitalization. Process mining is an important element in that journey.