For years, insurers have worked on improving customer experiences and on the necessary sales and marketing channels. Derived from this, the internal provision of services is also “available”.

This blog post addresses three central questions:

  • Why does the customer experience stand and fall with end-to-end processes?
  • Why should I be careful with RPA?
  • Are actbots the next logical step?

End-to-end processes

Improving automation and the customer experience starts with the process – not the technology. The end-to-end perspective has developed into a leading principle, because only a comprehensive end-to-end process organization ensures that all processes in a business process work together in the best possible way and towards the common purpose, namely the fulfillment of customer needs, optimized and coordinated. The concrete and measurable improvements concern the variable and fixed costs, quality aspects, speed and ultimately also profitability.

With a view to digitization, companies are required to face the opportunities and challenges and develop suitable approaches in order to maximize the potential benefits of digitization. The orientation towards end-to-end processes has proven to be a suitable approach to consistently interlink digitization into the processes.

Many insurers still have problems with a customer-centric end-to-end process alignment – even insurers who have already introduced new technologies such as RPA. You leave the necessary efficiency potential with an impact on the customer experience. More on this now.

RPA, chatbot and AI

(1) Robotic Process Automation (RPA) shows its strengths in clearly structured, repetitive and rule-based processes and tasks that are carried out by humans. Often RPA is used as a “quick fix” so that the larger process problems are not fixed. Without the right knowledge, insurers often do not know which process should be automated from the customer’s perspective.

There is no doubt that RPA can simplify insurance processes and reduce operational costs. However, because RPA cannot learn like an AI tool, insurers need to understand the nuances and downstream implications of automating each process so they can program their RPA solution to do the right things.

(2) A chatbot is a text-based dialogue system that allows chatting with a technical system. With increasing computer performance, a chatbot system has recently been able to access ever more extensive data sets and therefore also offer intelligent dialogues for the user. Such systems are also known as virtual personal assistants. Here are two examples:

  • Risky, an insurance bot, was developed by the fintech riskine. It includes pre-trained use cases, powerful speech recognition, as well as numerous relevant product information.
  • The chatbot Lizzy, an HDI bot, was developed on the AI-platform of e-bot7. The challenge was to select an ideal use case for lead generation and at the same time to enable users to complete the entire purchase process including payment via the PayPal integration – end-to-end. Today, HDI offers its customers the option of adding temporary drivers to their respective insurance policies as part of their motor insurance programs – fully automated in less than 3 minutes!

The interaction of the two technologies is interesting, in which chatbots are used as a front end for RPA bots. To start an automated process, the user starts a chat with a chatbot, which – during the conversation with the user – collects all the information required to have a task performed using RPA. If the user gives unclear answers, the chatbot has the ability to ask whether he has understood everything correctly before transmitting the information to the RPA bot. The RPA bot then starts the process and, if necessary, returns information about the chatbot to the user.

(3) I have already written several times about artificial intelligence (AI) at this point. Therefore, here are just a few selected use cases at insurers that were developed by AI specialists such as Leftship One, Layer7AI, datasolut and Automation Hero:

  • Churn Prediction – Identify high-risk customers with machine learning: Companies that constantly evaluate how customers use products and encourage customers to exchange opinions in order to resolve problems immediately are more likely to maintain a mutually beneficial customer relationship. For this purpose, customer data is collected in order to identify behavioral patterns, segment endangered customers and take appropriate measures to regain their trust.
  • Next Best Offer (NBO) – Machine learning provides the basis for NBO marketing. Historical customer data is used to predict future events. For this purpose, machine learning models are created that recognize certain patterns in historical data. It is thus possible to forecast what the customer is likely to want to buy next, for example which product or in which range is being bought.
  • Named Entity Recognition (NER) – the automatic identification and classification of proper names. In general, use cases for NER can be found wherever, for example, customers have the opportunity to enter free texts. Specifically: If an insurance company allows the customer to report claims digitally, NER can be used to pre-analyze these reports. Examples of entities that can be recognized here are parties involved in a traffic accident.

Process automation using Actbots

Intelligent process automation (IPA) connects the above approaches and tools with one another. Actbots are created, another form of digital assistant that will relieve people in the future. With contextual information processing, direct information exchange and fully automated processes, the new form of Actbots not only relieves the workload of employees in the company, but also offers customers and suppliers real added value thanks to higher service quality. In this way, queries can be automated more efficiently, processes streamlined and productivity increased. In the end, the company and the customer benefit from faster processing and a much better customer experience!

Conclusion

The insurance world is changing. Trends from the areas of smart analytics, industrialization and customer interaction, which together form the “digital triangle”, are finding their way into the insurance industry and are fundamentally changing products and processes of insurers.

The next level of process automation using Actbots is still in its infancy. For insurers, the possibilities seem endless – from the use of virtual assistants and agents, to chatbots for a novel interaction experience, to the detection of patterns and suspicious activities for fraud detection.