More successful new customer acquisition through automated broker offer process

Author: Dario Leuchtmann

One of the most important pillars of insurance is the acquisition of new customers. Without new customers, the business cannot grow. Brokers play a significant role in this for many insurance companies.

Brokers, however, are not committed to just one insurance company, but have contracts with many different insurers and recommend to their clients the offer that best suits them. If an insurance company processes requests for quotes from brokers, the probability is therefore high that the competition will come into play after all. This can be the case due to various factors: 

• Because the offer was simply sent to the broker too late.
• Because the complex process in the background at the insurance company is too expensive and the offer is therefore no longer competitive.
• Because information has been lost in the process and the offer made does not meet the client's desired requirements.

All these cases have the same core of the problem in common and therefore also the same solution. When the process is lengthy, complex and inefficient, it causes delays, costs unnecessary money and misunderstandings and errors occur more quickly and frequently. The solution to this is a simple, fast and clearly defined digital process. This gives employees a clear guard rail in manual tasks and takes a lot of manual work off their hands through automation and fewer media breaks.

Challenges: A lot of manual work and tacit knowledge

Brokers, in their role as external insurance advisors, do not have access to the insurance company's internal systems. This leads to the following challenges:

1. The quote requests come in unstructured in the form of an email.
   This creates a lot of work on the part of the insurance company to sort and evaluate the requests. Furthermore, not all brokers work the same way. Some make a bundled request for different products, others send a single request for each product. Keeping track of everything and sorting it all requires great expertise from the clerks.
2. Processing the enquiries and calculating the premium requires different know-how from several departments.
   Some of the clerks can handle the simple enquiries themselves, while the more complex cases are handed over to the underwriting department (UW). Since the calculation of offers for corporate clients can be very complex, there are different UW departments. The internal communication channels thus sometimes become confusing and therefore prone to errors.
3. The broker should receive a complete response to the quote request as quickly as possible.
   Service contracts exist between brokers and insurance companies, which regulate the maximum time they have to wait for an answer. However, clients increasingly want to receive their offer in a short time. It is therefore advantageous for an insurance company if the offers can be sent as quickly as possible. For this, simple, direct and fast processes are needed in which work can be done in parallel while still maintaining an overview.

Solution: New, modularly expandable process

In order to address the challenges mentioned above, we have developed and implemented the following business process together with a large Swiss insurance company.

1. Automatic recognition of the e-mail
   As shown in Figure 2, the offer process is preceded by an artificial intelligence (AI) that analyses the incoming emails from the brokers. In the process, the two most important key values, customer number and broker number, are read out automatically. These two values together allow a unique allocation of the request. Additional values could extend the automation of the process even more. However, it is advisable to concentrate on these key values first. This way, one can go productive faster and then train the AI further with the productive cases.
2. Automated offer process without media breaks
   The Business Process Management process (BPM process) is started automatically for the case handler without him having to look at the e-mails himself. All the necessary information is already available in the offer process. Any queries to the broker can also be made directly from the process. Thanks to the automatic e-mail processing, he receives the answer directly back into his already running process. During the user tests, we found that the employee can quickly become overwhelmed by the automated updates that arrive. Therefore, we designed the GUI in such a way that the user always sees at first glance what new information has arrived and, in particular, where it came from. This way, it is always clear to the employee where he or she is within the process.
3. Central overview for the clerks
   Our newly modelled offer process bundles the following three sub-processes for the clerks, as shown in Figure 2:
   
   1. overall overview of the status of the various quotation requests
   2. direct communication from the process with the broker
   3. direct communication from the process with the UW.
   
   This means that no more requests can be forgotten. Furthermore, communication between the various contact persons is accelerated, since the lead and the overview are with one person and the communication channels are shortened.

Cost/Efficiency

The knowledge about the process flow is modelled in a BPM process and is no longer only available as expert knowledge among the employees. As a result, fewer errors occur and the process runs faster.

Revenue

Due to the more efficient process flow, the offer reaches the broker more quickly. As brokers do not have to wait long for offers, the chance of a successful contract conclusion is increased.

Customer Value

The customer receives the offer faster, which increases his satisfaction and the likelihood that he will remain loyal to the insurance company.

Conclusion

With this solution, the broker insert process is greatly simplified and thus accelerated for the clerks. Existing distribution channels are thus still used and result in increased revenue.
During implementation, care was taken to ensure that everything is modular, so that individual modules can be easily expanded or replaced. Once the entire process has been tested and the employees have become familiar with it, it would be possible to offer an API interface to the customer app in addition to the AI for email recognition without much effort. This modularity should definitely be taken into account. Because it is precisely through such extensions as an API interface to the customer app that further considerable added value can be generated very quickly and cheaply in the future. The customer could, for example, independently make quotation requests for which he would receive an answer within a very short time.