- DIGITAL SOURCING
The solution aims to reduce claim leakage by analyzing historical claims data. A risk-based approach for monitoring of claims can be adopted to help improve the quality and consistency of case reserving. The solution empowers the client to drive its future policy decisions based on market intelligence.
Customer segmentation and profiling is imperative to drive marketing strategies. Our solution aims to identify micro-clusters that are actionable and specific strategies may be put in place for high impact clusters. The analysis is done by leveraging key parameters e.g. transaction data, demographics and psychographic information to develop profiles of customers across segments and value hierarchies based on micro-clusters. This leads to improved response rates and ROI by personalized marketing based on the customer's profile. Identification of profiles likely to churn can also be recommended.
Further, it is critical to assess the profitability of individual customers. This can help drive customer level strategies such as securing highly profitable customers from competitors and conceding permanent loss customers to competitors. Customers are classified based on their profile & buying behaviour. Customer profitability is derived by analyzing the product portfolio and claims history of customers
Our solution focuses on building strategies for cross-selling and up-selling of products, customer segmentation and customer acquisition based on customer profitability, customer lifetime value analysis, affinity analysis and sentiment analysis
The objective is to analyse and evaluate the contribution of each campaign in terms of its sales impact & ROI. In many cases, all conversion attribution is given to a single touch-point, typically the last customer touch point. However, a single source attribution is simplistic and multi-source attribution may be needed. A probabilistic attribution approach can be used to factor in multi-source attribution. Multivariate regression models are used to understand how campaign activities interact to drive purchases. Analysis related to baseline v/s lift, cannibalization, halo/drag effect and pre & post promotion effects can be done to derive the effectiveness of various campaigns.
The objective is to track a prospect's journey towards conversion across channels. The solution enables profile matching and highlights probable profile matches across channels with a specified degree of confidence. Business rules can be incorporated to facilitate customer mapping and a rules dictionary can thus be built up and enhanced over time. A feedback loop is also created in case a mapping was found to be incorrect.
The objective is to arrest the customer churn during the renewal process. In many cases, the customer may want to renew but the renewal notice does not reach the customer on-time or through the right channel, resulting in churn.The solution approach calculates the propensity of customers for doing a renewal based on his own historical behaviour or of the behaviour of the action cluster that the customer has been mapped to. Further, historical data mining helps in identifying the channel affinity for individual customers/customer clusters. Customers at risk can be approached proactively for renewal using the right channel, thus preventing the loss of customers.
A significant number of the agents are initially very productive. After a certain time duration, the agents drop out. The solution aims to predict and arrest such drop outs. Historical data is mined to understand agent profiles & associated behaviour and determine profile clusters exhibiting similar behaviour. Clustering of Agents is done based on demographics, educational, professional and performance parameters. Correlation between incentive pay-outs and retention is also assessed. A new agent coming in can be then mapped to an existing cluster to predict future behaviour.
The solution aims to assess the risk profile of the book of business and understand the various risk pools. This helps in deciding whether to acquire a risky customer. The probability of claims/claim severity from a customer is predicted based on the customer's demographic profile and historical claim behaviour. The forecast is then used to assess the risk profile of the customer. This helps in assessing the health of the book of business based on perceived risks from various customers.
The solution aims to identify probable frauds and raise a proactive alert. Internal as well as external data sources are considered and the associated data is primarily unstructured. Fraud identification can be made by applying advanced text mining algorithms & machine learning techniques on structured and unstructured data, backed by strong business rules.
The solution aims to improve turn around times for various processes such as new policy issue, cancellation, policy renewal, etc. which in turn leads to improved customer satisfaction.Turn around time expectations can be specified based on products & transaction types and alerts can be generated proactively for cases in which the turn around time is expected to deviate (so that proactive steps may be taken). For cases that do not meet the expectation, traffic lighting can be done based on the observed delay. Further deep dive is possible for cases with delay to identify the process step causing the delay.