Artificial intelligence (AI), intelligent automation, machine learning and algorithms are all buzzwords generating a great deal of hyperbole across a wide range of industries, markets and applications. The truth is that many businesses already have some form of intelligent algorithms in place supplementing the human decision makers.
Our business is a digital one, delivering products and services to upwards of 10 million customers. Interactions through our distribution model, our sales processes and our service capability generate vast amounts of data reflecting customer behaviours, preferences and changing demands.
We deploy robotic process automation (RPA), machine learning (ML) and natural language processing (NLP) in several crucial parts of our business. ML is prevalent in our pricing and data analytics and insight functions, and enables us to continuously improve the experience we offer to our customers.
Some of our machine learning algorithms identify patterns of suspicious customer behaviour, preventing fraudulent activity across our product base. As a result of this analysis we are able to prevent everything from fraudulent policies being taken out to misuse of our customer loyalty programmes.
The key is selecting the right ones for your customers and your business and to think ‘augmentation’ as much as ‘replacement’
Automated intelligent services extend directly to supporting our customers across our sales and service journeys. Virtual Assistants are deployed on our sales journeys, triggered by rules aligned to customer behaviours and designed to deliver help, information or education to our customers as they progress through a sales journey.
Our Virtual Assistant ‘Ask Becky’ pops up when the algorithm identifies a customer who might require assistance and offers to guide them onward in their journey. Our biggest group of products is focused around general insurance: home, car and van insurances. These are journeys which customers usually encounter annually, so quick questions and points of clarification which can be served in a conversational style online in real time make it easier for customers to complete the process in a single session.
Applications like these are considered as a ‘hygiene’ factor for digital products and services, so what does truly innovative look like? China has an aggressive strategy to dominate the world of AI by 2030. I took a small group of our progressive technologists and product owners there earlier this year to look at everything from niche start-ups to corporate and systemic applications of AI, which are driving value, commerce and consumerism across the region.
The first thing that hits you in China is the pervasive use of cameras capturing images as you travel around towns and cities. Leading in the field of Facial Recognition applications, China is developing, deploying and training these applications to drive access to services. This is most clearly illustrated in the Robot Hotel, which checks you in using facial recognition. The lift recognises you, stopping at the floor you need and your room door opens only for you! The whole system is integrated with Chinese identity system data, so checking in if you are not a Chinese national is a little more conventional, but it is still possible to be served at the bar by the robot bartender and the in-room tech does not disappoint.
Could this be extended to access to buildings such as schools or hospitals to protect people from those with less than good reasons to be there, possibly—or to business premises such as our own? Physical security is an integral element of our cyber and information security regime and this application lends itself to protecting our sites and assets.
Having access to vast amounts of data, and the power to predict, forecast and analyse patterns has resulted in an exponential level of rapid prototyping of new product propositions. Chinese companies have generated a wide range of micro insurance opportunities. Small value services are being identified, offered and accepted in moments and have a very short policy length, but of course, scale this across China’s vast population and it becomes very compelling. Protecting goods in transit, offering assurance of delivery, and even gamification of flight delays are all generated from highly specific models driven by huge data, machine learning and predictive models, with all this being pushed to the hand-held device of the customer.
In China, it is all about placing your product or service in front of the consumer to maximise convenience. Customers there are adopting the ‘superapp’ approach where a few brands (Baidu, Alibaba (AliPay), TenCent (WeChat), JD) ‘own’ the smartphone screen real estate and provide the customer with most of what they need to run their daily lives. It’s all about convenience for the customer. Providers integrate their products on the superapp, creating a seamless experience and placing the product into the hand of the consumer rather than multiple brands all vying for the customer’s attention, often forcing the multiple keying of personal details and a generally clunky, unintegrated experience.
My team has embarked on a programme of automating services provided through the voice channel. We are digitising voice services, using Natural Language Processes, models and automation to deliver customer service. After a number of significant strategic investments in recent years to complement our online sales capability, including a comprehensive service hub, we were faced with the challenge that customer preference was still to call us, partly as a result of our historical approach to providing phone-based services, but also as a result of the low frequency contact rates that annual insurance products typically generate. We took the challenge to switch the contact rate from the majority being handled by live agents to the majority being delivered through an automated solution without seeing a deterioration in the customer experience.
Whilst it’s early days, we are trialling leading-edge software, integrated technology and the art of Customer Conversation Design to put effective, customer-orientated service automation at the heart of our offering, whilst ensuring that customers who need live help from our expert customer service colleagues receive it quickly. This generation of voice recognition, NLP and intent modelling is performing very well and we fully intend to continue our investment in this space and adopt the very latest AI and conversational computing platforms so we can bring better and better experiences to our customers.
This technology is emerging quickly and it brings with it new challenges and a huge range of applications and opportunities. The key is picking the right use cases, leveraging the deep knowledge, talent and expertise we have in our people and getting something compelling in front of the customer as soon as possible. The science of AI, used effectively, opens up a world of possibilities. The key is selecting the right ones for your customers and your business and to think ‘augmentation’ as much as ‘replacement’.