The artificial intelligence (AI) team at ERGO has a small, highly efficient sister: Every two weeks, the colleagues in the Robotic Process Automation (RPA) unit launch new robots that quickly increase process efficiency in claims or customer management. RPA relies on structured - rather than unstructured, as AI does - data and processes. But what is particularly important to Mark Klein, the Chief Digital Officer, is that the robots have contributed more to the acceptance of digitization among employees than any other change measure. Four reasons for this.
For more than four years, I have been responsible for digital at ERGO as Chief Digital Officer. Before that, I held leading positions in the telecommunications industry. So I've already had some experience in the transformation environment. One of them is that digitization still worries many people: Some people fear for their jobs. They fear being replaced. My role therefore also includes translating, building bridges. Communication is an important part of my work.
For some time now, I have been supported by a technology that is not nearly as fascinating as AI and not as cool as some Alexa Skill. RPA - Robotic Process Automation - involves robots that, for example, fill Excel tables with customer data taken from various databases. These robots then carry out calculations in order to write the letter to the customer immediately afterwards.
Of all things, this pure process automation almost triggers a storm of enthusiasm at ERGO. The bots create acceptance for digitization. Concern has turned into curiosity, rejection into a real demand for the digital helpers. In this blog post, I explain the most important reasons that have contributed to that "cultural" success.
Each bot at ERGO not only has a personnel number created in the HR department - because only registered employees are allowed to access sensitive customer data. In some teams, the robots also have names. They are called Roberta, Probotti, Manuel Bü or T-Rex. This is one of the most important reasons for acceptance: A bot works like a clerk, merely mimicking the clerk's activities. You can literally watch it perform its work.
Imagine looking at your computer screen and watching a movie in fast-forward mode: an Excel spreadsheet is opened, data is copied, spreadsheet is closed again, another spreadsheet is opened, then a Word document, data is copied from left to right. 30 process steps that run before your eyes at breakneck speed. At the end, you see the letter to the customer being pushed into the outbox.
Afterwards, your eyes may hurt, but you were able to watch how the digital works. With a classic AI, it's different. You don't see (and understand) how image recognition fishes photos that include, say, railroad trains out of millions of photos. You don't know how the AI recognizes that that's a train in the photo and not a wall or a candy bar in the landscape. And, frankly, Data Scientists often really don't know much more, either.
That's a clear victory for the robots against machine learning! This kind of traceability creates trust. You know what the machines are doing. The mystique that often surrounds the digital is thus lost.
Typically, what makes your life easier grows on you. To me, that's the case with my insurance scanning app for submitting medical bills, and it's the same in my case with Google Maps. Even with the bots, colleagues quickly realized that the digital helpers provide real assistance with daily routine work. Effects such as "less overtime" were directly noticeable.
The robots take over tedious, monotonous, low-value-added tasks with high transaction volumes. Typing tasks that nobody likes to do. Open one system, copy, open the other system, and so on. With large insurance groups that have merged several insurers under one roof over the years - each bringing its own technical system landscape - this is not uncommon. That's why there's such a need for bots to link the systems together and do the mindless copy/paste.
For employees, this simultaneously means they can (and must) devote more attention to the more complex activities of customer and claims service. This realization also increases the acceptance that the robot relieves employees of repetitive processes in order to bring in their strengths such as creativity and emotionality in the sense of customer orientation.
This realization led to a fundamental rethinking in the specialist teams: With the prospect of remedial action, the willingness to critically scrutinize and automate other processes grew as well.
The approach to building the bots differs from classic large-scale IT projects: At RPA, we make products, not projects. Producing tangible results in a short time without changing existing system landscapes - that's what it's all about.
As a result, only a few weeks, rather than months, pass from the start of the project to the day a bot goes live. Some people talk about “no code - low code” here - in other words, a manageable programming effort. We work with Blue Prism, by the way. There are few technologies where such a short time elapses between plan and finished product. Two robots go live at ERGO Germany every month.
For our colleagues in those departments the bots are meant for, this is crucial. They have to pursue their actual job in customer service. There, capacities for projects are severely limited. The short sprint instead of a long project marathon raises motivation, leads to a better product and more acceptance for the technology right from the start.
The beauty is: RPA is capable of increasing employee satisfaction as well as improving the service experience of our customers and working more cost-efficiently. The technology has also been tried and tested for a long time and is therefore mature.
The teams work agilely according to DevOps principles: The classic separation between “one writes software (Dev), the other runs it (Ops)” is eliminated. Four to six colleagues form a team, composed in the roles of Process Owner, Subject Matter Expert, Process Analyst and Developer. They work in close coordination with the IT infrastructure.
The RPA team does not see itself as a full-service provider: Although it supplies complete solutions from a single source - from potential and process analysis to development, testing, coordination with the works councils and operation - the departments for which a robot is being built are deeply involved in the project.
This cross-functional work is another principle we promote at ERGO: don't just digitize and put it in front of the target department. Instead, involve those specialist colleagues and develop things together with them. Make the bot their bot!
But RPA - and this will come as no surprise - is not something we have set up as a “cultural change” program. We are using the technology to achieve business goals: improve customer experience, simplify processes, leverage efficiency through automation. With this KPI, we are on target as well.
ERGO Germany alone now has 60 bots in use. These process more than 60,000 transactions per month. For certain “business transactions” - as we say in the nicest insurance jargon - the processing speed of customer concerns has increased significantly. The pre-selection of cases by the automatic machines, which have to be checked more thoroughly (by a real human), is also getting better and better.
But I admit - I prefer to talk about the qualitative parameters, the feedback from specialists and managers, board members and works councils. The best sign of acceptance is the sharp increase in demand. When the RPA team started out, colleagues in the specialist departments went to the trouble of finding suitable processes for the bots in the first place. In the meantime, the processes are coming to them.
In this way, RPA has become a real door opener for other technologies such as data analytics, artificial intelligence and voice assistants. For a Chief Digital Officer, that's like winning the lottery!
Text: Mark Klein
Most popular