SAP Process Mining or how to understand your business processes

Look at this picture. In the business processes of many companies the same situation - a lot of workarounds that can simplify the actions of employees, while not violating the accepted rules. But how can managers see this reality inside the company?

The easiest way is to contact the consultants. What does the consulting market offer today to solve this problem? Conduct interviews with employees and find out how they subjectively see their processes. And they can see it as the left path in the picture. Or they may not even know what this path should be. The problem appears: how to see the objective reality, and not the interpretation of someone's point of view? This applies to any business processes - procurement, sales, logistics, etc.
')
There is a separate SAP Process Mining solution in the SAP product line that allows you to see the “digital steps” of company employees. In fact, one can see any process from any information system, track the compliance of employees' actions with corporate policies and standards, and various reference models. The scale of the company does not matter here - the main thing is that the process itself be digitized, and not carried on paper.



For example, in the figure above, a user starts a transaction in the process of creating a sales order. He creates a purchase order, makes a shipment from the warehouse, then sends it to the customer, gives him an invoice, receives payment. Please note that at the moment when the user conducts transactions in the system, a transaction log appears in it. It shows that a certain user at a certain point in time created an order No., changed the price and delivery address, changed such and such documents, and so on.

And such logs can be collected from any systems, including samopnyh, up to text files and excel. The SAP Process Mining solution has its own internal algorithms that can collect logs and turn them into graphs. These are pictorial pictures (see Figure 3) that show how the process actually looks without anyone's interpretation.


Fig. 3

As a result, you can see the bottlenecks on each project. For example, orders piled up somewhere. The manager cannot process them quickly. The question is - why? Either there is not enough resources, or the process itself is methodologically built so that you have to spend a lot of time.

In Figure 4, we see a graph of the process “Processing a loan application”. Each red ball is a credit application. The size of the ball is determined by the cost of the loan application (the amount claimed by the applicant). Looking at the graph, we see that between some activities there are queues of balls - lines of red color. This suggests that at this point in time, applications accumulate between stages or are processed for a long time. The reasons may be different: lack of resources, personnel, features of the regulation processes, staff qualifications and other reasons.


Fig. four

By the way, bottlenecks are well illustrated in the procurement. Especially in a multi-branch holding, where this process may be different in each division. Such processes are formed historically - there can be different information systems, processes are carried out according to different methodologies. In this situation, sooner or later, the parent company will want to bring them into a single template in order to make it easier to control. The task of Process Mining is to show the difference between the affiliate process and the benchmark, and also what needs to be done to bring the processes in the branches to a single standard. Figure 5 shows an example of the comparison of two branches: on the right — the reference, on the left — to be changed.


Pic.5

Process Mining is applicable for robotization tasks. With this method of process management, you can see where many routine tasks exist and think about how they can be robotized.

For example, a company often adjusts prices for orders from certain suppliers and for certain groups of materials. And how much time do employees spend to adjust prices for 145 thousand orders a year, each of which takes 40 minutes? How much time and resources can be saved if this amount of work is reduced by at least 30%? How much will the duration of the procurement cycle decrease if you robotize this process? As a result, people will be freed up to solve analytical and managerial tasks. Figure 6 shows a table with a list of suppliers for which a large number of manual adjustments are made.


Pic.6

Process Mining allows you to "catch" and fraud scenarios. For example, in a company, a certain sales territory is assigned to each manager, with stores in which you need to sell goods. At the level of the NSI system, this relationship is fixed and then transmitted to the CRM-system and reporting system.

What happens at the end of the year? The seller enters this system and detaches several stores in which he failed to sell the goods. And it turns out that his sales plan is now fulfilled not by 80%, but by 100%. He is paid a premium, because he made his plan. Time passes, the new year begins, and he re-adds himself to these two stores. After conducting an internal investigation using Process Mining, it turns out that managers and their colleagues had access rights to the system, the ability to change data and make adjustments. An example of a rights and permissions management report is shown in the figure below.


Fig.7

In the following articles about Process Mining, we will talk about sales and receivables scenarios, analyze application examples in the procurement and payment processes, and with one of our partners we will do a detailed overview of the integration and data upload.

Authors: Ekaterina Tyuleneva, SAP Business Solutions Architect, and Fyodor Pavlov, SAP Platform Solutions Expert.