Hyperpersonalization in Insurance Is No Longer Hyperbole
Insurers are finally waking up to the need for a proper data modernization strategy after the COVID-19 rude wakeup call.
Many chief digital officers and chief data officers are busy tidying up their master data management and investing in analytics capabilities. While they may be setting up the building blocks for driving better customer relationships, the real goal is hyperpersonalization.
It is a term whose definition is as murky as its vaunted goal. Yet, analysts say hyperpersonalization will be a game-changer. It is leading the likes of IDC to boldly predict that 15% of customer experience applications will deliver hyperpersonalization through reinforcement learning algorithms continuously trained on a wide range of data and innovations.
What on earth is hyperpersonalization, really?
Let’s start with that definition, which is better explained with an example.
Personalization is now the basic expectation of customers. So, getting a birthday card, sending a calendar with your name, or giving some form of personalized acknowledgment shows that you care about them.
Hyperpersonalization takes a step further by understanding their needs and wants at the moment and responding to them in real-time. In insurance, this is gold as it allows you to sell policies and riders at a time when the end customer is ready for it.
The problem with hyperpersonalization is that it is resource-intensive. And before, scaling it to all your customers may have been difficult. But in today’s AI-driven and data-rich world, this is no longer the case.
In its “How hyperpersonalization can drive customer growth post-pandemic” article, EY noted insurers are investing in new data sources, analytics platforms, and AI-powered decision engines as a result. It allows them to match customers with solutions while allowing them to follow each customer based on the moments.
However, like anything in data science, you cannot just switch on hyperpersonalization. And you need to address three major concerns that have less to do with the technology itself.
Unraveling data politics
Moving and merging data pipelines that used to end at interdepartmental data stores is never trivial.
Ask any chief data officer, and he, she or they will tell you that you need political grit and a senior management mandate. If you need the data to be shared on a modern data infrastructure like a data lake, you need a strong carrot and stick.
COVID-19 provided both. It slowed business but also heightened awareness about personal protection. This saw many insurers looking for better ways to monetize their customer relationships. Meanwhile, digital-savvy insurers and insurtechs are beginning to erode the market share of the incumbents.
The biggest challenge for CDOs is trust. Many managers and agency heads, steeped in old ways of selling, see AI-based systems as challenging their market wisdom. Getting them aligned is challenging because, unlike banks, agencies are run like small businesses.
“Hyperpersonalization depends on a deep understanding of the customer, which needs systems and applications coming together to create a single source of truth of reliable customer data to deliver relevance and personalized approach,” says Kenneth Koh, head of industry consulting at SAS Asia Pacific.
This requires the agency and middle managers to participate in the data projects. Koh notes that once you integrate online and offline data to create a 360 view of the customer, you can drive data trust. Data governance and ethics reinforce data trust further.
Another roadblock is the agency workforce itself. It sees data-driven hyperpersonalization as a human workforce replacement strategy.
“But this whole thing about data analytics, it's not meant to replace the agents,” he explains.
Instead, Koh urges companies to market hyperpersonalization as complementary. “And if you look at it from a broader perspective, the foundation layer of hyperpersonalization is going to lead you towards the omnichannel capabilities and where the agency workforce is but one channel.”
Koh also advises insurers to go beyond the first-party data that MDM projects focus on. Adding social media data and other third-party data, like telecom data and soon IoT data will allow them to derive contextual insights quickly — which is the whole point about hyperpersonalization.
Becoming a customer-centric organization
Secondly, for hyperpersonalization to work, it needs both top-down and grassroots support. And often, this requires organizational change.
In fact, Koh highlights that the one reason hyperpersonalization faces resistance is that the insurer is still organized along product lines. “You need to become customer-centric and not remain product-centric.”
This is difficult to do. While many insurers have been paying lip service to customer-centricity for years, they still think along product lines. Just ask for the best quote, and you’ll be given product recommendations with expected premiums.
It is not just underwriting that faces challenges when it comes to hyperpersonalization. EY, in their paper, noted that marketing in insurance is “highly decentralized” and differs from country to country. It is one reason why there is a difference in experience between visiting the corporate website, talking to an agent, and engaging a distributor website.
“Each part of the organization is focused on the next-best action (NBA) for the customer based on what’s narrowly best for their area of purview — not what’s ultimately best for the customer,” wrote Avril Castagnetta, EY’s Americas insurance marketing transformation leader and Bhuvan Thakur, managing director for FSO technology consulting.
So, to maximize hyperpersonalization, the entire organization needs to change its approach toward customers. “Firstly, the marketing team probably needs to change the way in terms of how they want to operate. Secondly, we need to find ways to engage with agents as in Asia, we’re still very much face-to-face.”
Koh notes that insurers need to do some soul searching. They will probably reorient the company towards customers and shift from a product-centric P&L to a customer-centric one, said EY in their paper.
Koh believes that hyperpersonalization may be the answer to drive this mindset shift. For example, insurers can quickly explore new relationships and create bundled offerings with the correct data and analytics. And with the increased customer engagement (via bundled offerings), insurers can offer better contextualized data and lead to more relevant recommendations.
Don’t forget the outliers
Lastly, we need to change the way we handle analytics. We often try to normalize the data. But when it comes to hyperpersonalization, outliers matter.
“Outliers are important. And in fact, it could be a moment of truth,” says Koh.
For example, if I buy a single policy for someone, after all the years of spending on myself, I may be seen as an outlier. Yet, attentive insurers might see this as a new life moment and maybe a change in life stage.
“What an insurance company can do is to quickly extract this information and identify who I bought the policy for. If I bought it for my family members, let's say a child, then likely there is a change in my life stage,” describes Koh.
The difference that AI-driven hyperpersonalization brings is that you can do this at scale and even determine the subsequent actions an insurer needs to take — all potentially in real-time. This opens new cross-sell revenue opportunities that current efforts miss.
“In fact, with predictive modeling, it can let us know whether you will be interested in buying another policy. So that's the difference. Do not discard outliers because they may reveal certain patterns or information which we may have taken for granted,” says Koh.
Such information will be crucial as the insurance industry braces itself for a new yet lucrative battlefield: wealth management. And here, hyperpersonalization will determine the winners and losers.
Winston Thomas is the editor-in-chief of CDOTrends, HR&DigitalTrends and DataOpsTrends. He is always curious about all things digital, including new digital business models, the widening impact of AI/ML, unproven singularity theories, proven data science success stories, lurking cybersecurity dangers, and reimagining the digital experience. You can reach him at [email protected].
Image credit: iStockphoto/Jirsak
