“Has the insurance industry encountered its Amazon moment”? Well, opinions may differ on this subject. Early adopters of Automation and Artificial Intelligence may have divergent opinions on the business benefits delivered through efficiency. Before we go on to debate further on this topic, I would like to start off by narrating two examples from my recent experiences.
I ordered online from a retail store, which offers pick-up and delivery from nearby physical outlets on a Sunday. The next day, Monday, was a public holiday. The expectation is that the order will be processed on the same day. To my surprise, I learnt on Tuesday morning that since the last two days were holidays, the back-office team would process the order when they get back to office. This made me wonder if this was indeed a 24x7 digital experience.
In the second example, I happened to be involved in the process of preparing for a hurricane at a mid-tier insurance company. The operations team was grappling to arrange enough staff and makeshift customer care centres to respond to the spike in claim intimations that were expected. Now, it is pivotal to understand that the claims process is a decisive moment for a customer, and defines experience in a big way. Responsiveness during such catastrophic events goes a long way to define customer experience.
These are just two examples that makes us look at AI and Automation from a very different lens than purely in terms of cost savings. The intent of this article is to discuss the challenges faced by incumbent insurers in the wake of changing demographics, customer behaviour and increasing natural calamities - and how Intelligent Automation can be explored as a solution.
What are some of challenges facing traditional insurers?
24x7 support – Changing expectations of customers and field sales representatives leading to demand for anytime, anywhere service models.
Responsiveness, consistency and quality - Pressure on reducing the time to respond for requests from customers and agents/ brokers
Data as a competitive lever - Data collected through the digital footprint of customers provides more data points for personalization, underwriting and claims assessment.
On demand scaling – Events such as natural calamities create a spike in volumes of work inflow, requiring flexibility in size of the workforce.
Efficiency – Traditional insurance companies operate at 35-40% expense ratios, of which 8-12% pertain to operational expenses.
Time to market – The legacy application ecosystem is an impediment for rapid rollout to new markets.
In summary, traditional insurers are faced with transforming their business models, operations and IT to stay competitive in a changing landscape.
Strategic choice for insurers?
Most traditional insurance companies operate through an intermediated distribution model, and manpower intensive front and back-office operations, supported by legacy systems hosted on an on-premise infrastructure. In order to stay competitive, companies resort to one or all of the following initiatives.
- Improve efficiency of business and IT operations by modernizing legacy applications or migrating to a state-of-the-art core system (core system refers to policy administration, billing, claims and reinsurance).
- Building a digital engagement layer comprising portals and mobile apps, integrated with legacy systems through a services layer.
- Achieve incremental efficiency in a short to medium term scenario by adopting a combination of Robotic Process Automation (RPA) ,Cognitive Automation and AI tools
Are these initiatives mutually exclusive? In our opinion, the answer is NO. Each of these initiatives have similar objectives. The quantum of benefits and the timeframe to deliver them, the cost of the initiative and risk of failure are some of the considerations for organizations to sequence these initiatives.
The focus of this article is to further explore Intelligent Automation, which provides faster and incremental benefits to a business without too much disruption to their current operations.
What is Intelligent Automation?
The insurance industry is back-office intensive and given the legacy ecosystem, a sizeable human effort is involved in running the operations. The scope of automation can be broadly categorized as follows in order of complexity and adoption.
Traditionally, automation is targeted at repeatable back-office tasks acting on structured information. Activities such as reading customer service requests logged in a workflow system, reading requests coming through in a structured format such as submissions in a pre-defined template, capturing the information in back-office systems, handling exceptions and taking corrective action by sending a mail when mandatory information is missing etc. This is Wave 1 of automation, where the scope was limited to the above described activities. This wave is completely dominated by RPA tools. While there may be limitations on applicability, we have witnessed improvements at 40-50% efficiency in processes where feasible and applied.
The next wave in automation is to target the customer, and intermediary interactions and inputs that come as unstructured information. On analysis of the contact centre statistics, 60-70% of calls that come in are related to enquiries. A good proportion of these enquiries are simple, which can be automated with a partially human-like approach. The answer is chatbots and voice bots which will not only replace human agents, but complement them for simple conversations and out of office hours support. This application can be extended to e-mail channels as well, where they can be interpreted, categorized, assigned and when feasible, automatically responded to. Insurance, being a customer-centric industry, care has to be taken that this does not come in the way of human empathy that is required in an engagement. If anything, they should enable human agents to spend more time with customers for more engaged conversations as well as bring in a faster response and 24x7 support to top end customers. Natural Language capabilities can also be utilized in interpreting unstructured information such as non-standard data submissions coming from brokers and data enrichment by collecting key pieces of information from third-party data sources.
The third wave in automation is the utilization of Deep Learning and Machine Learning capabilities. Machine Learning can enable accuracy of prediction. For example, we see Machine Learning being used to improve the accuracy of conversations in chatbots. Additionally, Computer Vision and Deep Learning are used in the retail industry and can find their applications in the inspection and claims process.
A visualization of how Intelligent Automation works:
Let us explain this with a simple example – John Smith and his family plan a weekend trip to Chicago from their residence in Minneapolis. It is Friday evening, and John realizes that he has not added his son Geoff as a named driver in his insurance. The call centres are closed for day. He realizes that he has received an e-mail pertaining to the launch of a bot to assist customers after office hours and weekends. He opens his Skype messenger and starts a conversation to add his son to the insurance. The interaction is intuitive and bot assists him in collecting the necessary details and uploading pictures of the driving license. The information collected is processed by the back-office bot, which provides him with an additional premium. On his consent, the payment is processed from his existing credit card on records. John is delighted as he could get this simple endorsement processed without having to wait till Monday or call his agent/ broker.
Thus, Intelligent Automation brings together Robotic Automation, Conversational Solutions, Natural Language Processing, Character/ Image Recognition techniques, third-party data extraction and enrichment tools to bring in a holistic approach to automation. The below diagram depicts the key blocks of Intelligent Automation.
Why Intelligent Automation?
We have witnessed the following benefits delivered by Intelligent Automation:
- For a personal lines insurer in the US, automation of back-office processes has resulted in close to 50% improvement in productivity, thus enabling them to focus on customer engagement.
- For an Australian insurer, the processing time for claim payments was reduced by half of what it was before.
- For a major US insurer, simple conversations were enabled through chat and voice bot. This resulted in a 40-50% reduction in call volumes for conversations.
In addition to the above, overall, it has resulted in improved customer experience, quality, consistency and responsiveness.
What should organizations do?
Insurers need to evaluate the choices before them and the timeframe each of them take to deliver business benefits. Insurers need to look at Intelligent Automation as a short to medium term solution. Companies that have either embarked or are yet to embark on Digital and / or core transformation, can utilize Intelligent Automation to free up resources as well as achieve immediate results without overhauling their applications and infrastructure.
Mindtree enables its customers on their automation journey by utilizing a structured methodology for defining a roadmap from the current to target state. This will encompass identifying suitable processes, interventions required for holistic automation, sequencing initiatives by business functions and determining the business case from these initiatives.
To know more about Intelligent Automation for the Insurance Industry and how automation is changing the insurance industry landscape do click the link