Success factors in the hyperautomation of processes
The term hyperautomation stands for the next evolutionary stage in the automation of business processes. With the support of artificial intelligence, partially or fully automated processes become self-optimizing workflows or even important digital advisors for process managers. In an interview with DOK Magazine, Johannes Keim, Partner at Ventum Consulting, explains what it takes to make hyperautomation a success.
Question: What goals are companies pursuing with hyperautomation?
Johannes Keim: Basically, hyperautomation is the next step in the digitization of companies. Digitized processes become fully automated, self-optimizing workflows. Several specific goals can be pursued in the process: Companies want to achieve sustainable growth and improved efficiency through the appropriate AI-supported automation of processes. Often, this is also intended to counter the consequences of the shortage of skilled workers. It is also about becoming a more attractive employer and gaining a competitive edge. But successfully implementing hyperautomation is not trivial. On the contrary, it is important to proceed in a structured manner. The first step is to select particularly suitable and promising processes for hyperautomation.
Question: Technologically, there is always talk of artificial intelligence in connection with hyperautomation. How important is AI?
Johannes Keim: AI already has a very central role. It is the enabler for this form of automation. After all, AI today allows us to automate tasks where this was not even conceivable a few years ago. AI-supported, for example, signatures of a person can be matched with each other fully automatically – for fast and very reliable authentication. There are also a wide variety of other repetitive activities that can be hyper-automated via AI, at all levels of the enterprise. Or think of chatbots – another example of automation with AI. Today, chatbot AI is often so well trained that it is no longer easy to tell whether you are chatting with a real person or a robot.
Generally speaking, hyperautomation generates added value in the context of a wide variety of customer-centric processes – for both external and internal customers of a company. In addition to AI, however, there is another technological component that adds momentum to the topic. These are the low or no-code platforms. They empower departments to program by themselves, significantly lowering the barrier to automating processes.
Question: With which processes should a strategy for hyperautomation best start?
Johannes Keim: It always starts with the business case, and calculating it is not that difficult in principle. The rule: The investment in the automation of a business process must be covered by the savings generated within a defined period of time. That is why it is often advisable to prioritize processes with high throughput. Examples would be account openings in the financial industry, customer inquiries in the service sector or processes in fields such as data structuring and governance.
With suitable software solutions, it is even possible to create an almost completely hyperautomated line of business. Consider, for example, the use of self-services with AI-based document, email, telephony and chat routing. Based on process throughputs, advertising can then also be switched on automatically, or cloud resources can be added or reduced as needed. In principle, the potential benefits of hyperautomation are always great when decisions are made according to clear rules and patterns.
Question: Which path leads to success in hyperautomation? Do you recommend the big bang or an incremental approach?
Johannes Keim: In the marketing texts of the players involved, hyperautomation is often described as an immediately available panacea for process inefficiency and overworked employees. But it is not. If only because the path to successful hyperautomation has a lot to do with whether the organization already has a digital DNA. Otherwise, the company would be faced with a huge transformation task until hyperautomation could be deployed across the board and economically. It is therefore often advisable to initially develop only certain, carefully selected processes at a manageable cost.
If automation projects fail after all, or at least take significantly longer than planned at the beginning, there are reasons for this. Either the hyperautomation strategy has prioritized the wrong business processes, or the business requirements do not have the level of detail needed for the project. In addition, hyperautomation requires a willingness to completely reengineer business processes. And the company’s database needs the level of maturity required to make AI analytics-based hyperautomation even possible. Conventional reporting standards are not enough.
Question: Apart from the business case, what other criteria are there for prioritizing which processes are best to start with in your hyperautomation strategy?
Johannes Keim: Hyperautomation becomes particularly rewarding when it addresses functionalities that are relevant for numerous processes of a company and that are reused. This is how you get the required impact on project goals. It is usually a good idea to build the backbone of the business model first. Take call center processes, for example. Here, in the area of telephony, telephone routing and the stable connection to the telephone provider would be possible initial areas of application. Only then will a well thought-out hyperautomation strategy address services that are more complex and more specialized – these will then have a significant impact on the project goals.
In the call center example, a worthwhile next project would be the AI-supported, automatic identification of customers based on their voice. If the automatic voice recognition works, further services can then be implemented: for example, the visualization of the contact history or an AI-supported determination of the next best action. These automation steps result in call center agents increasing their first resolution rate. The ultimate goal of hyperautomation in this context would be to avoid the call altogether – for example, by proactively offering self-services with the help of pattern recognition or by completely automating the Next Best Action, including the corresponding conclusion.
Question: That already sounds like extensive digitization. But you are still in favor of proceeding in well-considered individual steps?
Johannes Keim: Yes, absolutely. It also remains sensible to follow the Minimal Viable Product approach and initially implement only the absolutely necessary functional requirements. Automating well means doing less – and doing the right thing. As companies move toward broader hyperautomation, it’s critical that they prioritize the right things and start small. First, it is important to determine the maturity levels of business requirements, data, IT architecture and organization. This is the only way to determine the processes for which hyperautomation is already possible and promising. In most cases and industries, the transformation of the company can only take place step by step. You also have to realize that this issue is not about a mere tool. Ultimately, hyperautomation is a vision of the company’s future; it has strategic relevance. In the ongoing digital evolution of enterprises, it is the next big step. This cannot be achieved from one day to the next.
Partner at Ventum Consulting and expert for process automation