Process Mining in banks – an important element of digital transformation
Leonid Chernyavsky, Development Director of RAMAX Group, spoke about how Process Mining technology helps Russian banks (and not only) optimize their operations, how to use PM even more effectively, and what to expect from the technology in the near future.
We restore the real course of business processes on digital traces.
Let's start with a small educational program. What is Process Mining, what is it?
Process Mining is a set of methods and technologies necessary for conducting high-quality digital transformation of any business processes. Process Mining analyzes the digital footprints of users in various information systems and builds a map of events or activities within the company's divisions in real time, indicating important actions that do not comply with regulations, as well as deviations from the execution of the process.
In every large company, on the one hand, there is a relatively clear description of the sequence of actions of the staff, recorded in the regulations, instructions and protocols. On the other hand, there is also our own understanding of the company's management, based on some operational analytical reports from classic BI-systems. On the third side, there are employees of the company who every day with their hands perform part of the operations of a large end-to-end business process described in the regulations or present in the BI report on the manager's desk. These employees have absolutely their own understanding of the specifics and flow of each specific sequence of actions. But when Process Mining, a kind of universal X-ray machine, is introduced, the real picture of the flow of certain operations is" highlighted "by the participants, their actions, deviations from regulations and instructions," bottlenecks", then it turns out a completely different story – the fourth and closest to reality vision of the organization of business processes in the company. The traceable digital footprints that users leave in the systems are the most reliable data. Experience shows that 95% of processes can be improved. The Process Mining technology is able to demonstrate this clearly.
As a result, we get the following. The company has a standard set of tools that help collect and analyze information from corporate databases. These include the above-mentioned classical systems of operational and strategic business analysis (Business Intelligence, abbreviated BI), which provide reporting in the format of dashboards, graphs, and tables. On its basis, top managers can track profitable and unprofitable business lines, as well as the dynamics of income and expenses, the sales department can monitor the implementation of plans and optimize planning, financiers can plan budgets, receive consolidated reports and manage expenses. These systems do a good job of evaluating KPIs, modeling the results of actions based on the decisions made.
If we talk about Process Mining, then these methods and technologies allow us to look at a set of different metrics, assess how many options there were for executing the process, see what deviations there were from the basic implementation path, and how long it took for a particular action. With all these tasks, the classic BI is not able to cope.
At the same time, it is important to understand that any business process is a whole chain of events, which implies a different sequence of actions for its individual participants. For their analysis, the Process Mining class systems are just right – they contain mathematical algorithms that allow you to quickly and efficiently restore the real course of business processes using digital traces.
The financial sector is a leader in the implementation of Process Mining
Over the past few years, Process Mining in Russia has been most actively implemented by the financial, oil and gas and telecom industries, but banks are the undisputed leader among them. Most of the business processes there are already digitized in one way or another, which creates the most favorable conditions for the use of Process Mining. In principle, in financial organizations, there are two groups of business processes where the technology is most actively used.
The first group is all front-end business processes related to customer service. The second group is back-office processes. In 2016-2019, RAMAX Group implemented a number of joint projects with VTB 24 Bank and VTB Group. Our fruitful cooperation continues in the difficult 2020.
We started our joint project activities by optimizing the issue of loans to individuals at VTB24 Bank. The key to the success of this project was the combination of several factors. First, banks are always digitized and have been working digitally for a long time, and VTB24 is no exception. We quickly found processes where Process Mining can be successfully applied. Secondly, there are quite progressive people working in banks, so the human factor played into our hands in this case. Undoubtedly, we must pay tribute to the people who worked and promoted this story in the bank itself. Third, we took the right business process for analysis and optimization, because in 2016 there was a relatively good macroeconomic situation, and the bank's retail unit was actively developing. He understood that the growth in the number of loans issued to individuals should be stimulated, improve the quality of customer service, and engage in the classic capture of market share.
As a result, it was these three factors that positively influenced us to successfully launch the project. Next, we switched to lending to SMEs, and later switched to salary projects. As for the second option of using Process Mining, processes for the back-office, VTB technology touched on the procurement process.
As a result of the implementation of the project, due to the increase in employee productivity, the productivity of processes increased by 6-8%. The evaluation of the sales technology used made it possible to replace it with a more efficient one and reduce sales losses by 3.5%. Labor costs were reduced by 8.5% due to optimization of extreme process scenarios.
After the public announcement of the successful implementation of the project in VTB24, all the largest banks in the country began to actively show interest in the technology, and today there is no bank from the TOP 20 that we would not contact about this. We have launched similar projects in the field of Process Mining with a number of them.
Among the public ones, we can mention the case with the UBRiR bank. By implementing Process Mining, we provided the bank with transparent metrics for both the entire business process and its individual sections. In the four months since the implementation of the project, the bank has received its first results: sales of credit products have increased by 4%, and an action plan has been outlined, the implementation of which will allow the bank's customers to spend 20% less time in the branch when processing services, which will significantly speed up the business processes of the UBRiR retail unit.
Related technologies
As already mentioned, for the successful application of Process Mining, it is necessary to analyze the processes that took place in the past. Sometimes parts of processes are registered in different information systems, or one employee conducts different processes in different systems. If there are any (approx. - systems) in the same banks, we are faced with the fact that we get data for analysis with different levels of detail for logging user actions.
The overall picture for the sequence under study is collected in two ways. The first is to expand the possibility of logging the analyzed system, provided that this does not limit its performance. The second is the installation of special software (so-called "agents") on the computers of users who participate in a particular process. This software captures any user activity on their PC, which allows you to see a detailed cross-section of operations in a particular process. Let's say a person opened a browser, and the system recorded it. The information is collected from a personal computer and transmitted to a centralized database, from where we can consume information about digital footprints. Such programs greatly enrich the information stored in other systems.
Among the technologies that can work in conjunction with Process Mining, it is worth mentioning RPA (robotic process automation – from the English. robotic process automation). Process Mining helps you track which operations require manual execution and predict the effectiveness of their robotization. In some industrial enterprises, the utilization rate of robots reaches 50%, but PM allows you to achieve even greater efficiency.
Another technology that deserves attention in the context is machine learning. Some training models are already "embedded" in Process Mining. Such a "bundle" helps to observe the already standardized process, identify its "bottlenecks" and undesirable deviations from the regulations. The training of models is based on the company's past experience, which helps to make predictions for future processes.
Development of Process Mining in the Russian market
In Russia, the number of successfully implemented projects in the field of Process Mining is still small relative to world statistics. No more than ten successful implementations can be found in the public domain. Most of all in the Russian market, large brands from the financial industry and the manufacturing sector need PM. They get the fastest possible economic effects from optimizing customer service processes, improve its quality, reduce costs and production cycles, and increase the efficiency of the process of maintenance and repair of equipment. As for the process areas, Process Mining is most often used to optimize procurement and logistics processes, for example, in industrial enterprises or in the retail industry.
Among foreign vendors in the Russian market of solutions, Celonis is the leader. However, domestic developments have also begun to appear, which we actively monitor, and we contribute to the development of some of them.
The relatively low growth rate of the number of projects for the implementation of Process Mining is due to the fact that now many companies in the Russian market are not yet ready to implement such solutions. This is often due to the fact that some industries have not yet completely gone "digital", and many processes are not yet sufficiently automated. At the same time, there are positive trends. For example, the trend towards the digitalization of the public sector allows us to predict the growth of corporate interest in technology.
Trends in the application of Process Mining. Not a single sequential process
At the end of 2019, all Process Mining implementations followed a roughly identical scenario. We considered the business process to be optimized (either a complete end-to-end process, or some part of it), and improved its efficiency. Then we moved on to the next process. And so, gradually and as far as the customer's resources are sufficient, we went from one process to another, developing an internal competence center that supports the Process Mining technology within the bank.
Today, we can distinguish two more trends where it is appropriate and effective to use Process Mining. The first is a scenario where there are two identical processes and they need to be combined or "straightened". This happens when, for example, one company has bought another. Both have a procurement department, and you need to smoothly transfer one company to the processes of the other, which are recognized as reference. Celonis Process Mining helps this process to pass most painlessly.
Another trend is when a new system is being introduced, for example, ERP. With the help of Process Mining, the migration process from the old system to the new one is 30% faster and much less expensive. We can already see "digitally" how the previously installed solution worked and what the new process looks like using the new system, how the old processes migrate to the new system, whether they follow the planned path or the path is also somehow transformed. The main value here is that the person responsible for migration sees the whole picture immediately digitized online and (a key factor) can manage it.
The future of Process Mining – simulating business processes and creating digital counterparts of banks.
I will allow myself an allegory by imagining an organization, in our case, a bank, as a living single biological organism. Everybody has its own organs, there is a circulatory system. We look at the business process from the point of view of the circulatory system that affects all organs, and improve it, thinking that the result will "spread" to all organs sooner or later, and a single positive synergy effect will be achieved.
But if you look at a particular process in particular, it does not work by itself, it somehow affects all the other "bodies" of our company. So, it is important that it is possible to look at the totality of the work of all the organs-processes in the company and predict how changes in one will indirectly affect the others. As a result, we get a system where we consistently see the influence of some factors on others. It does not always happen, but all processes are somehow correlated with each other and affect each other.
RAMAX Group has launched the SIMBA product line, which is an evolution of Process Mining technology for simulating business process scenarios. The solution is the next step towards building a digital transformation environment and creating your own Digital Twin of Organization (DTO). With the help of SIMBA, any bank can not only optimize one or several business processes, but also build a digital environment to increase business profitability.
The digital double provides a business model of the bank's functioning, simulating the reaction to changes in the external business environment or internal processes. This reduces the time and cost of implementing organizational changes and risks associated with digital transformation, and eliminates the need for experimentation. SIMBA from RAMAX Group allows you to set the parameters of the planned changes in efficiency (step duration, time spread, change in the course of the process), get an automatically calculated result of changes, taking into account all the nuances, and see which indicators most affect the efficiency of the bank.
But this is already a topic for a separate column.
By materials PLUSworld.ru