Insurance Solutions

Analytics value across the value chain

We help insurers attract, manage and retain profitable customers by efficiently and effectively applying a wide range of advanced analytics capabilities tailored to each business need.

Pincer Pricing Optimization
CLTV
Inspection of Risk
Agency Management

Pincer Pricing Optimizationgo to top

Business Problem: Insurers frequently change rates to bring prices in line with the losses experienced among various customer segments. Customers react to rate changes in different ways, such as through attrition and rate avoidance. Most insurers have limited capabilities to predict and quantify customers' collective reaction to rate changes and their long term impacts. Having this capability allows insurers to perform accurate forecasting and planning, as well as demand-based pricing to maximize the benefits of a rate change.

Solution: We build price elasticity models for insurers, using historical data, to predict customers' reaction to price changes. We also build lifetime value models to predict the long term revenue potential of customers. The Pincer framework allows insurers to combine their traditional loss models with price elasticity models and lifetime value models to create a powerful pricing framework. With Pincer, insurers can forecast the short term and long term impact of rate changes. This approach also enables the tailoring of rate changes to achieve business goals.

Deliverables:

• Price elasticity model
• Lifetime value model
• Loss model (optional)
• Pincer framework to enable insurers to use the models effectively for price changes
• Recommendations and support to enable insurers to maintain and support ongoing use of the Pincer platform

CLTVgo to top

Business Problem: From insurers' perspective, each customer has a different value. This value is the total revenue/profit that a customer is expected to generate for the insurer over the length of the relationship. Insurers can use the lifetime value measure effectively across multiple touch-points such as marketing, claims handling, policy servicing, etc. Today, most insurers lack a solid framework to quantify the lifetime value of each customer.

Solution: We build lifetime value models for insurers, using historical data, to predict the expected longevity and cross-sell potential of each customer. Insurers use these models to design their customer relationship management strategy around the lifetime value framework. For example, customers with high lifetime value can be given preferential treatment when they call the customer service center, whereas low lifetime value customers may be deferred to an automated response system.

Deliverables:

• Lifetime expectancy model
• Cross-sell potential model
• Framework to regularly score all customers using the LTV framework
• Recommendations to use the LTV framework effectively

Inspection of Riskgo to top

Business Problem: Insurers incur significant cost to inspect the risk they want to insure. Risk is inspected either internally, or via a third party data provider, allowing the insurer to accurately price the risk. This expense can be reduced substantially if insurers can identify high risk segments where inspections can be avoided. Currently, most insurers do not have a sold framework to identify such segments.

Solution: We build predictive models for insurers, using historical data, to estimate the outcome of a particular risk inspection using available information. These models can be used to identify segments that are much less likely to have a negative outcome from the inspection of risk. An ROI framework may also be developed to maximize the returns on risk inspection expense, in addition to enabling a reduction of the related expenses.

Deliverables:

• Model to predict the outcome of inspection of risk
• ROI framework to enable insurers to design new strategies for inspection of risk
• Recommendations to enable insurers to use the framework effectively

Agency Managementgo to top

Business Problem: Agents are a critical driver of insurance business success, however profitability varies by agent. Insurers need to carefully distinguish unprofitable agents due to inferior business parctices from those who are unprofitable due to the inherent randomness in claims. Currently, insurers lack the comprehensive data-driven and systematic scorecard approach they need to rank agents based on performance, and manage them accordingly.

Solution: Our solution consists of developing two models: one at the policy level and one at the agent level. The policy-level model is designed to predict losses as a function of policy characteristics. The agent-level model is designed to predict the agent's profitability as a function of its business practices and policy characteristics. The agent-level model allows insurers to rank agents based on their expected profitability, allowing them to reward profitable agents and take corrective measures against agents generating higher than expected loss.

Deliverables:

• Model to predict policy-level losses
• Model to predict agent-level profitability
• Agent scorecard to rank agents
• Recommendations to use the agent scorecard effectively