The Insurance industry works on predictions based on customer profile and statistics to come up with insurable events and to calculate premiums. While this works well for insurers, customers may not always get the best deal. The solution? Delivering personalized quotes by leveraging the power of data analytics. Accenture estimates that 88% of insurance consumers seek personalized offers, messages, pricing and recommendations from their auto, home or life insurance provider.
But building up a personalized profile to deliver quotes is easier said than done.
Insurance and banking companies are sitting on a goldmine of customer data residing in silos that is being underutilized. The problem is that the disparate systems in these companies might interoperate to an extent but are not necessarily interconnected. Add to this, the data in each of these systems are changing on a regular basis.
Challenge #1: It is important that these disparate systems need to be effectively integrated to make a unified view of the customer.
Insurers already use transaction and demographic data extensively to segment insurers. A vehicle insurer creates a quote based on the type and colour of car, the location (whether the car is parked in a garage), safety features etc. But such data is not enough to drive hyper-personalization.
Personalization requires data drilled down to the individual level. In a recent survey, The Realities of Personalization Report by Econsultancy & Monetate, while 94% of the insurance companies agree that personalization is critical to their current and future growth, almost half say that IT roadblocks (47%) and legacy technology (46%) are “major barriers” to their personalization efforts.
Challenge #2: The key to delivering a personalized experience lies not just in the effective integration of multiple systems, but by deriving actionable insights out of the data residing in them.
A quick look at the data in these multiple systems show that apart from the existing systems, the data for personalization is obtained from open sources such as open.gov, software trackers, and third-party data vendors. Data is also collected from voluntary subscribers who may fix a box in their car, and transmit data via a wearable or IoT sensors. Several insurers and insuretech firms like metromile have already started using such data to offer personalized advice and quotes.
Such live data coming from IoT sensors allow insurers to track asset stability and performance on real time basis. Insurers may subject such live stream data to analytics, unearth anomalies from the norm, and offer timely warning to the customers to take pre-emptive action. Such initiatives create a better engagement with customers, and may aid to revise the premium during renewals. For instance, data from a black box installed in the car allow insurers to give personalized driving advice, and may also be useful to get an accurate
picture of the risk profile, for renewal. But getting hold of such data the first time round is extremely difficult, if not impossible.
Using personal data, regardless of the source, also carries the risk of privacy violations. Moreover, the costs of acquiring such data and reconciling it to the enterprise requirements can quickly add up. If the acquisition cost is too high or exceeds the lifetime premium itself, the exercise becomes self-defeating.
Challenge #3: The data from different sources, subject to privacy concerns and industry regulations, need to be sanitized and then reconciled with business objectives, packaged into actionable formats and finally integrated seamlessly into analytical models and enterprise systems.
What is the solution to creating a personalized quote?
The insurance industry is doing its best to get a handle on these challenges using a mix of technology and clever marketing nudges. For instance, insurers are using Artificial Intelligence to make predictions on the individual, based on whatever data is available. However, AI is only as good as the algorithm which powers it. It would require a highly mature AI algorithm, based by adequate data in the first place, for an AI system to deliver hyper-personalization. There is also the risk of the AI algorithms carrying over the bias of its makers.
Many insurers develop behavioural nudges through their marketing campaigns. They identify some common ground with the customer, and reframe their messages to make them more involved, so they volunteer more information about themselves. One popular tactic is offer multiple and different engagement options, to help identify preferences.
While insurers are on the spree to take their customer experience to a next level, a basic foundation must be in place to make it practical. Such a foundation can only be built with a customer data platform in place. Mining the rhythms of customers with the available data and predicting their needs before they even realize it, is the future. Data science and Machine Learning models built on top of a customer data platform, such as Suyati’s Buyer Rhythms (BR), can help meet these challenges.
About BR: Suyati’s Buyer Rhythms engine is an AI powered advanced analytics built on top of a Customer Data Platform that will help insurers optimize their marketing and create hyper-personalized customer experiences. It also enables rapid data driven business iterations which will lead to strategic decision making. Know more.