RPA Example: Financial services provider MLP automates its processes with Servicetrace

As a forward-looking German financial services company, the MLP Group had been considering the deployment of robotic process automation (RPA). Company management quickly recognized the many benefits: including lower costs through faster, error-free processes, less administration, and higher employee and customer satisfaction. MLP selected the German RPA solution, XceleratorOne (X1) from Servicetrace as its preferred technology and the company introduced robot-controlled process automation in just 14 months. MLP also installed an effective controlling model for RPA.

The MLP Group has been the financial services partner of choice since 1971 for more than 555,000 families and 22,400 business customers. Across Germany, 1,800 employees and more than 2,000 consultants manage assets of more than 48 billion euros. The topics of digitalization and robotics were triggered internally by company management – they wanted to know: What opportunities can robotics offer the company? What can RPA do and what are the benefits of this technology? The company assessed the automation of business processes and quickly realized that there were significant opportunities and benefits to be gained from the deployment of software robots, process automation and artificial intelligence.

One of the drivers for robotics and automation was the technological situation at MLP Banking AG. As one of the “non-major banks”, MLP Banking AG relied on an external core banking solution, which meant a lot of duplicate data entry and which, at that time, was not sufficiently integrated into other MLP systems. The result was an impact on data quality and time-consuming manual work processes, especially in the area of accounts & securities. In MLP’s banking area, therefore, the demand for automation and digitalization, including RPA, was especially high.

How can we automate processes? From the idea to proof of concept

To get started, MLP assembled an interdisciplinary team, which collected information and compared the products of established providers. This selection committee then developed a catalog of criteria for selection of the technology. The start of 2020 saw the first feasibility tests with two providers: During the proof of concept (PoC) phases, which ran over several days, the objective was to successfully automate two selected examples and thus to check the feasibility, the software, and the internal framework conditions. Due to the first corona lockdown in the spring of 2020, the PoCs had to be carried out virtually. After this reality check, MLP opted for the German RPA provider, Servicetrace and its RPA software, XceleratorOne (X1).

“The comparison of two providers in a PoC was extremely helpful for us. We favored a different provider to begin with but, ultimately, Servicetrace and its unique X1 platform were clearly superior to the competition”,

Timo Niethammer, Project Manager Organizational Development at MLP with responsibility for RPA.

“Automation with the Servicetrace X1 platform has produced excellent results. We were also very impressed with the intelligent image recognition technology, which is essential for automation in a Citrix environment like ours. The Servicetrace image recognition technology worked better than the Citrix plug-ins from other providers. In addition, we were very happy with the high level of customer focus of the Servicetrace Consulting and Support Team, the chemistry between them and the MLP staff was right from the first minute”, says Niethammer.

Reasons for MLP to opt for the Servicetrace solution:

  • Simplicity: The intuitive and easy application of the X1 platform promotes acceptance among users in the specialist departments. After all, the MLP approach is to enable business users for automation.
  • Structure: The Servicetrace integrated lifecycle methodology enables a company-wide, standardized, and very structured approach.
  • Compliance: The lifecycle approach of X1 is optimally suited for audits and checks in the highly regulated finance sector since the platform makes introduction and implementation fully transparent and verifiable.
  • Image recognition: The intelligent and exceptionally high-quality Servicetrace image recognition is significantly better than that of the competition and it enables and simplifies process automation with Citrix.
  • Customer support & Consulting: The Servicetrace in-house consulting and customer-support team was MLP’s immediate and competent contact partner during the initial introductory phases.

Company-wide rollout in three phases in just 14 months

In the period from January to May 2020, the MLP selection committee evaluated and selected both provider and product, carried out the PoC, and made the decision to purchase. Due to the corona outbreak, the initial phase of the introduction took place entirely virtually. From June until October, extensive preparations and further planning were carried out: The team wanted to develop an operating model for automation by the time the necessary IT architecture and infrastructure were set up. At the same time, Servicetrace began training of selected MLP employees who, in addition to their previous range of responsibilities, wanted to become part-time RPA designers. The final stage of the introduction involved a five-month rollout and knowledge-building phase. In this final phase, the primary task was to develop initial prototypes from the process pipeline and to transfer these to productive operation. This included, for example, the development of the Happy Path, the extension of this path to include error handling, processes relating to testing of the automation workflows, and launch and monitoring of the first process automations.

Topics and decisions in the preparation phase:

  • Operating model: MLP decision to opt for operation in the cloud.
  • Setting up the process pipeline: Collection of processes that are suitable for automation, in collaboration with the specialist departments.
  • Process evaluation: On one hand, the technical feasibility and, on the other, estimating the added value of automation of a process.  MLP uses a reference value of 80 working hours, which a process would need to save each year in order for the development effort to pay for itself after approximately one year
  • Work & communication structures: As they gain practical experience, the RPA designers become more efficient. MLP starts with an agile working approach and with deliberately short time periods and concrete goals for the design of the automation processes. The team reflects regularly and openly on the status, the results, and the experiences gained.

Company-wide introduction of process automation at MLP in three stages.

Company-wide controlling model with a decentralized Center of Excellence (CoE)

The core of the CoE consists of four employees, two of which have a representative function. Project manager Timo Niethammer was involved right from the start and is one of the active CoE members. He works almost full-time for the CoE and, in his role as project manager, he collaborates closely with the decentralized project participants in the business areas. For MLP, one factor was very clear during the implementation: The specialist areas are most familiar with their own processes. Therefore, the employees from those areas should be empowered as process designers for optimization and automation. Up to March 2021, this consisted of six persons from two specialist areas. In the subsequent months, three more specialist areas and their users followed. Responsibility for operation of the processes also remains within the business areas. This means that, if an automation process fails, the relevant alarm registers in the immediate business area so that the employees can react quickly.

The tasks of the CoE at MLP:

  • Development of an RPA strategy
  • Functional & technical operation of the RPA tool
  • Central control of the RPA
  • Bringing together of all relevant stakeholders
  • (Further) development of standards, best practices and processes
  • Service & support for the technology
  • Knowledge transfer to the organization
  • Management of change toward an automated organization
  • Reporting to managers and C level

Decision process for process automation:

  •  Employees submit ideas for process candidates to CoE
  • CoE works with employees to evaluate the potential and the feasibility (with process evaluation)
  • Decision on automation: No –> Feedback to specialist area
  • Yes –> Application owners are informed by CoE
  • Inclusion of the process in the process pipeline for an upcoming RPA camp.

Communication and collaboration are the key success factors

The operating model for process automation developed by MLP covers three elements in addition to the technical topics: 1. defined stakeholders, 2. the management bodies and 3. regular communication formats. The stakeholders include the executive board, the individual specialist areas, the CoE and IT department, and the process designers. The high-level management bodies for the company-wide automation consist of the CoE, the ‘mission control’ meetings, and a week-long RPA camp every month.

The mission control meetings monitor the progress of all process optimization measures including RPA. The CoE evaluates and prioritizes the process candidates in the pipeline in consultation with the business areas. Every three months, the process designers and the CoE look again at all the information and decide which processes they want to implement in the upcoming quarter. The key factor here is the prioritization in the process pipeline. With a period of three months, these sprints are unusually long since the designers are working on process automation in addition to their regular tasks. This longer cycle also takes different learning speeds into account. The designers of process automation work in teams of two, which promotes dialog and knowledge transfer. The tandem model also helps to ensure that, in the digital working world of corona, no one feels alone. This is also the reason why an open-plan office is simulated in the RPA camps. All designers work with an activated camera even though each person carries out their own work. This atmosphere promotes a sense of community and lowers the obstacles to interpersonal communication.

Regular close dialog with the IT department is also a fixed component of the CoE as the department is involved in every automation. On the one hand, for operation of the IT infrastructure, and on the other, because the robots, also referred to as bots, access IT-operated systems and applications when carrying out automation. Close collaboration pays off and raises the quality and efficiency of implementation.

Since the go-live in February 2021, MLP has successfully automated six processes and has created a pipeline for processes to be automated. With the processes automated so far relating to topics such as processing standing orders or creating securities accounts, MLP has reduced the effort required for manual activities by more than 12 hours daily.

Success factors of the RPA implementation at MLP:

  • IT for company-wide RPA deployment is essential
  • Small core team made up of IT & process management for coordination of activities
  • Management buy-in ensures provision of necessary resources
  • Balancing of work focus in the project and in daily business
  • Regular dialog and communication with the stakeholders are important
  • Set goals for short time periods, regularly review experiences, and optimize procedures
  • Working in tandem promotes learning and strengthens team spirit

The management bodies, stakeholders and communication formats for RPA implementation at MLP