Preference management refers to the data and analytics that are activated when a piece of customer communication goes out (to a policyholder, agent, or claimant) at the time and in the way that customer prefers to be contacted. These communication channels are both digital and analog—text, phone, email, direct mail—and include marketing, transactional, and service-oriented communications.
For insurers, preference management is especially important due to the sensitivity of much of this communication, such as underwriting approval, claims, or payments. These touchpoints with the customer can occur at stressful or emotional moments, and insurers need to understand the preferences of the customer as they seek to provide them with the optimal experience.
Preference Communication
Most insurers rely on the customer to provide that preference in some manner. There are effectively three types of preference: stated, inferred, and predicted. Stated preference is how preferences are most widely understood. It is the most straightforward type; the insurer simply asks the agent or policyholder how they want to be contacted, such as “Go paperless” or “How can we contact you?” as part of onboarding or regular self-service. Alternately, a customer might proactively request a communication channel, most typically through a “subscribe” or “unsubscribe” link. This data can be stored locally, as part of an email or SMS platform, or centrally in a CRM.
But relying on stated preferences can result in many customers showing “no preference,” simply because they haven’t stated one yet. By waiting for the customer, the insurer loses out on providing the desired communications experience. Inferred preference, on the other hand, means using behavioral data to analyze how a customer has historically engaged with different communications channels to deduce their preferred channels.
Agents who routinely use SMS to communicate with underwriters can be marked out from agents who mostly do so through email. Customers who have never opened an email communication can be removed from those lists and placed on “phone” registers. By doing this, insurers can learn from customer engagement methodically and continuously and start to personalize the overall brand experience.
But what if there isn’t enough historical data on which to base an inferred preference? Predicted preferences are based on look-alike or next-best-action models that gather everything known about a customer, either through first- or third-party data sources, and forecast what those preferences are likely to be. Rather than being based on that customer’s prior behavior, the recommendation is based on the wider customer population and how other, similar customers preferred to be contacted.
At some insurers, customer experience engineering platforms deploy AI in real time in order to recommend the optimal communications channel. Customer data platforms connected to marketing suites might integrate with third-party data in order to recommend email, direct mail, social, or other media channels as being optimal given a customer’s demographic, life stage, or financial situation. Predicted preferences are among the many insights that can be derived from this kind of customer analytics.
Preparing for All Preferences
When thinking about preference management, insurers usually focus on only stated preferences and ignore inferred or predicted preferences. But insurers need to think about these last two for several reasons:
- Stated preferences can be stored locally—as a separate field within a CCM or email marketing platform, for example. But when insurers start to predict or infer preferences using data analytics, they need to centralize the data in a CRM or other customer data platform. Adopting inferred or predicted preferences thus also requires customer data centralization.
- Inferring or predicting preferences also has implications for who owns preference management within the organization. The deployment of analytics and personalization requires that preference management ownership involve IT and data science teams. Preference management ceases to be about communications and becomes part of customer experience engineering.
- One cannot conduct analytics without data. Instead of passively recording stated preferences, insurers need to actively collect data about how customers are interacting with different communications channels in an omni-channel ecosystem. That means proactively setting up data standards and analytics tools as well as the organizational governance that requires.
Preference management is one piece of the customer journey orchestration and engineering capabilities that data analytics makes possible today. As insurers renew their focus on customer experience, successfully managing their customers’ communication preferences takes on increased importance, even if customers haven’t yet stated what those preferences are.
Add new comment