Across the world, the insurance industry struggles with a reputation issue. Customers feel distant from financial corporations as they are often believed to be self interested and opportunistic. Feelings are rarely positive, with skepticism, indifference and distrust being the most likely consumer attitudes towards their insurance company.
Because insurance touches so many parts of our lives- cars, homes, pets, travel, health- it is also highly complex. Insurance markets are typically very competitive so firms need to stand out and at the same time meaningfully build trust and improve retention whilst keeping costs down. Cracking the enigma of customer engagement is a vital building block to reaching these goals.
One strategy is to develop a lifetime partnership with the customer, building a constant relationship with a customer for the duration of their life. The concept takes the customer lifecycle as a literal process for engagement. This goes beyond just reacting to an enquiry or a complaint and focuses on prevention and assistance in all stages of the clients’ journey. Most insurance and financial services firms would like to be able to work on this but are prevented by day-to-day operations and ‘firefighting’ which drains budget, agent time and management attention.
Why should you read this case study?
If you are working in the financial and/or insurance sector and are responsible for the customer experience or customer insights within your company, then you should read and share this study. The case study gives a clear actionable example of how customer feedback analysis can help build a better experience for your customers as well as save make an impact on cost management.
Both B2B and B2C
In this project Wonderflow was asked to investigate the different communication channels the large international insurance company has with its major stakeholders:
- Business to business: the relationship between the company and partners and agents
- Business to consumer: contact with end-users
The opinions and feedback of both were collected on a multitude of touchpoints, including online reviews, social media, internal CRM systems, emails to the customer service, chats, surveys and many more.
We used this huge range of texts to configure our Natural Language Processing (NLP) engine, transforming customer language into business topics and rules. Using NLP takes our work beyond just text search- it allows a full understanding and extraction of the most important information from text and builds further machine training based on its own results. The AI also allows Wonderflow to generate sentiment analysis and present results according to business logic, making it actionable.
The challenges we faced
Wonderflow was asked was to find information on what customers want and when and find out how they want to interact with their agent/provider. However, we also faced a number of challenges:
- Technology: The client operated an old backend CRM system that had limited exchange options with modern marketing systems.
- Responsibility: no designated manager or expert had been appointed, making the digital turnaround hard to manage.
- CXO Buy-in: No or little support from the board and higher management.
Results #1: B2B – Agent Experience increased by 40%
Wonderflow analysis developed new insights to increase the commerciality of products to increase sales and to decrease the number of unnecessary contacts to the customer service to decrease costs.
In the first case, Wonderflow analyzed around 16,000 chat interactions between the company headquarters and the regional insurance agents to improve sales performance and knowing the specific needs of geographical areas. Actions taken included:
- Develop strategies to prioritize short term complaints from the agents
- Improve timeliness and quality of communication with agents
- Show how their feedback is handled and processed
Impact: The company increased agent satisfaction by 40% in 6 months. This has an impact on cost of acquisition of new agents and on commission structuring.
Insurance companies need to learn how to make decisions based on the voice of the customer
Results #2: B2C – Emails to customer service decreased with 8%
Wonderflow analyzed all emails received by the customer service team (around 20,000 per month) including both inquiries and complaints. We ran this through our technology, configuring our NLP engine to understand the texts and build sentiment analysis and categorize the kind of interactions agents had. We aligned this with corporate priorities through our dashboard to make results understandable and meaningful. The aim here was to anticipate common questions and issues in order to prevent them from happening while increasing customer satisfaction.
Actions taken based on the Wonderflow analysis:
- Re-configure communication on the consumer website
- Re-develop the customer service knowledge base with more Q&A
- Clarify documentation and term & conditions of the products
Impact: Using the insights presented by the Wonderflow dashboard, emails to the call center decreased by 8% in the first 3 months alone. This was achieved through implementing minor changes, such as changing the format of the accident report. This, in turn, gave the marketing department the opportunity to interact with the client on different, more positive topics.
The first step towards building a real relationship with the customer has been taken by transforming the nature of incoming calls and messages. Now, agents have the time and space to focus more on serving the customer and building trust rather than problem- and complaint-resolution.
The information in this case study is confidential. Brands and company names are anonymized.