When insurers look to the future, they see artificial intelligence (AI).
From autonomous vehicles that use computer vision to live translations made possible by artificial neural networks, artificial intelligence is maturing quickly.
Customers are at the heart of the industry’s accelerating AI transformation.
According to Accenture’s newly-released report, “Technology Vision for Insurance 2017,” AI is becoming the new user interface, and has the potential to make every UI simpler and smarter. In fact, one-third (32 percent) of the executives surveyed said that their own company would be “completely transformed” by AI within the next three years, and an additional 39 percent said it would “significantly change” their company. We are seeing that
The report notes that, to keep pace with this change and turn today’s technology disruption into their own long-term competitive advantage, insurers are investing significantly in AI, including embedded AI, machine learning, deep learning, natural language processing and computer vision. Almost four in five respondents (79 percent) believe that AI will revolutionize the way insurers gain information from, and interact with, their customers, with 55 percent of the insurance executives surveyed citing better data analysis and insight as key benefits of embedding AI into user interfaces.
Some insurers have already put AI to work as part of the intelligent automation of the back office. Many carriers are experimenting with or deploying automated data capture and recognition, robotic process automation (RPA) and cognitive computing as levers to increase operational efficiency, improve customer experiences and let their teams focus on value-adding activities.
An invaluable tool
From the automation of rules-based business processes, the next big step forward for AI is in the creation of interfaces that make complicated technologies and processes more approachable for customers, agents and employees.
However, we also found that insurers — in response to growing customer demand for a better digital experience — are increasingly deploying AI on the frontline of their business to empower their agents, brokers and employees in delivering an enhanced customer experience. Two-thirds (68 percent) of insurers, for example, are already using intelligent virtual assistants to provide their customers with a better overall experience.
AI is becoming an invaluable tool for insurance agents in offering their customers the best advice about the many and complex products they represent from multiple carriers. In the future, it will be commonplace for agents and bots to work in partnership to optimize a quote or provide financial advice, with AI mining multiple datasets to offer contextual insights into the customer’s needs, the products in the market and other factors.
Geico, for example, has launched a virtual assistant called Kate for its mobile app. Kate understands natural language and can answer basic questions, such as “What is the current balance on my auto insurance policy?” or “What is the date of my next payment?” Assistants like Kate will, in time, evolve to answer more difficult questions and support the sale of more complex insurance products.
The research also indicates that, within the next three years, insurance customers may spend more time engaged with AI than talking to company employees or agents. Customers will, according to the research, base their opinion and interest in a company on its AI, just as they judge by their experiences with human employees.
The rapid maturation of artificial intelligence is causing insurers to re-examine, not only how they process transactions, but how they communicate with their customers, employees and agents. (Photo: iStock)
Huge potential benefits
Instead of interacting with one person at a time (the way a human representative does) an AI system can interact with an infinite number of people at once. AI-supported relationships can exist and grow across interfaces and communications styles, taking the form of text-based chats, spoken conversations, gestures, or even virtual reality. People can choose how much and what kind of interaction they want to have with the company at any given time.
Not only can AI maintain a consistent brand experience through every interaction, it can also use learning capabilities to tailor that experience to each individual, and evolve the experience to react to any new product or strategy the enterprise wants to implement. What’s more, the research shows that many insurance customers are already prepared to seek and buy service and advice via AI.
The potential of AI for insurers is enormous. To bring that potential to life, however, insurers must redesign their existing systems to support its features and technical dependencies.
Insurers are already leveraging AI and machines to help their people do their jobs better. Many adjusters, for example, use Tractable’s deep learning systems to simplify the triage process after a car accident. Instead of manually scanning pictures, they use machine-trained estimates for repair costs, helping agents accelerate claims past the triage stage and into repair, salvage or appraisal.
The big news, however, is in AI’s move from the back end to the front end of digital insurance operations. The question for insurers then becomes: How ready are you to explore AI to its fullest potential? The research found, for example, that challenges in integrating AI into an insurer’s business are top of mind. Data quality, privacy and infrastructure compatibility were cited as the top three barriers.
Few insurers have developed AI capabilities within UX/UI teams, or have trained those teams to take advantage of existing AI toolkits. Even fewer have begun making the changes to business processes and infrastructure that will be necessary for AI implementation.
The tech work ahead
Insurers will need to develop connections between systems and interfaces, and then between different points of interaction. Legacy systems can often be incompatible, and at times, this problem is compounded by the fact that insurers are taking a step-by-step approach with the targeted — and limited — introduction of new technology platforms, rather than making a wholesale change. The combination of the two makes for a difficult equation.
What’s more, accomplishing this technology transformation requires robust sets of data from every channel, not just to train the AI to interact with customers and employees, but also to enable AI to continuously learn how those interactions should evolve over a lifetime. This explosion in data also brings with it serious concerns about data privacy.
With the emergence of AI, we are entering an era of disappearing technology, in which customers no longer need to understand complicated technology to use it effectively. They can simply talk to, gesture at, or touch the AI that controls it. As technology becomes invisible through AI, the barriers to its use disappear as well.
This is all happening at warp speed. We believe that, within five years, more than half of insurance customers will select a carrier based on its AI rather than its traditional brand. Within seven years, most interfaces will not have a screen and will be integrated into daily tasks. And, within 10 years, digital assistants will be so pervasive that they will keep employees productive on a 24/7 basis, operating in the background for workplace interactions like creating video summaries immediately after an important meeting.
“AI is the new UI,” meaning that artificial intelligence will be a central element in the way insurers design and implement their new user interfaces. (Photo: iStock)
Key implementation steps
Insurers need both short-term (next 100 days) and intermediate-term (next 365 days) plans to put AI into action.
Short term, such plans might include:
- Rank customer, agent and employee interactions by how critical they are to current revenue and future growth.
- Pick three to five interactions and working with service and product teams to streamline the way customers engage with your most valued products and services.
- Identify the information and insights that would help improve the user experience.
- Develop AI personas that fit your brand and communicate your brand voice.
- Consider piloting AI in roles like content curator (for personalization) or interaction advisor (for intelligent automation). Think about starting with an internal process such as IT support.
- Identify communication channels and platforms for integrating conversational experiences with your brand, considering both internal as well as external interactions.
- Gather and review existing key performance indicators (KPIs) for customer success. Ensure these KPIs account for the benefits of simplified interactions.
The 365-day plan might include:
- Implement increasingly sophisticated AI personas that not only curate or advise, but also aim to orchestrate as much as possible among your brand and ecosystem stakeholders for key customer interactions.
- Develop analytics that take account of front-end customer insights and back-end business intelligence to better understand key customer interactions.
- Develop and pilot a training program for AI teams and UX/UI teams to cross-train on implementing AI to improve and simplify key customer interactions.
- Design a new customer experience in which AI serves as your frontline brand ambassador for customer service interactions, communications, and engagement with customers.
The research shows that AI will become increasingly important to insurers’ success in the digital economy. When properly designed, AI can empower customers, agents, partners and employees, creating value at every customer touchpoint. Insurers who get AI right can build deeper customer engagement, stronger brand loyalty and more-sustainable profits.
John Cusano is Accenture’s senior managing director of Global Insurance.