Building the recommender for curated insurance bundles on a voluntary benefits platform

You are a product manager at an insurtech company. The company wants to meet the insurance needs of a corporate client's employees by launching a voluntary benefits platform. Rather than selling individual insurance products, the platform offers curated combinations of insurance plans. How would you go about developing the system that recommends these plan combinations? The question assumes a working knowledge of insurance and of employee benefits. Voluntary benefits are optional coverages that an employer makes available to its staff and that employees usually pay for themselves, typically through payroll deduction. ```hint Start from one employee Before choosing how the system decides, describe what a good recommendation would look like for one specific employee, and list what you would need to know about them to judge it. ``` ### Clarifying Questions - Does the system recommend combinations to individual employees at enrollment, to the employer deciding which combinations to offer, or both? - Is the insurtech the insurance carrier, or a distributor that packages plans from partner carriers? - What employee data may the system use: employer-provided records, answers employees give, health information? - What should the recommendations optimize: enrollment, coverage that fits each employee's needs, revenue, or retention at renewal? ### What a Strong Answer Covers - How voluntary benefits are offered and enrolled in, and who the stakeholders are - A definition of what makes a recommendation good, and how it balances fit against business goals - The inputs available, how they are gathered, and the privacy and regulatory limits on using them - A staged plan for building the recommender and the experience that presents it, including how recommendations are explained - Success metrics with guardrails against unsuitable sales, and how the system improves over time ### Follow-up Questions - How would you recommend combinations for a new employer whose workforce has no history on the platform? - If the recommender raises enrollment but many employees drop the coverage at renewal, how would you diagnose it? - How would you check that the recommendations are fair across ages, income levels and other groups?