Pricing Analysis with DataRobot at NTUC Income
Customer Company Size
Large Corporate
Region
- Asia
Country
- Singapore
Product
- DataRobot
Tech Stack
- Automated Machine Learning
Implementation Scale
- Enterprise-wide Deployment
Impact Metrics
- Cost Savings
- Productivity Improvements
Technology Category
- Analytics & Modeling - Predictive Analytics
Applicable Functions
- Business Operation
Use Cases
- Demand Planning & Forecasting
- Predictive Replenishment
Services
- Data Science Services
About The Customer
NTUC Income is a top composite insurer in Singapore, serving over two million customers with 3.7 million policies. It is part of the National Trades Union Congress, the sole national trade union center in Singapore comprised of 58 trade unions and 10 social enterprises. These social enterprises were established by the government to help stabilize the price of commodities and services, strengthen the purchasing power of workers, and to promote better labor-management relations. NTUC Income is the only insurance cooperative in Singapore, providing life, health, and general insurance products to over two million customers across the country. It is both the top composite insurer in Singapore, as well as the largest automobile insurer.
The Challenge
NTUC Income, a top composite insurer in Singapore, was facing rising claims costs across the insurance industry. As the cost of doing business increased, the company needed to understand the factors driving up claims costs, who was affected, and what actions to take. Furthermore, with insurance increasingly becoming a commodity, accurate price setting became more critical than ever. However, pricing analysis in insurance can be complex, repetitive, and time-consuming. The traditional method of using Generalized Linear Models (GLMs) for pricing analysis was not ideal due to several limitations. These included assumptions of a straight-line relationship between a rating factor and claim costs, time-consuming processes, and inability to analyze text in claim descriptions. The company needed a solution that could address their pricing analysis challenges and scale with their team.
The Solution
NTUC Income turned to DataRobot, an automated machine learning platform, to undertake a massive pricing analysis project. The platform was used to identify changes in exposure, claim frequency, and the severity and nature of claims. DataRobot's Feature Impact and Word Cloud capabilities were used to identify patterns in claim frequency and visualize how claim descriptions were changing over time. The insights gained from the data allowed the actuaries at NTUC to set more accurate technical pricing. They could get a better estimate of the risk premium, identify mispriced exposures, and account for future inflation. The final step was to transition from technical pricing to commercial pricing. Using DataRobot, actuaries could analyze a sample of competitor quotes and generalize out to see what competitors would price for different types of policies. This allowed them to balance profit margin vs. volume and find a practical premium rate that customers would pay.
Operational Impact
Case Study missing?
Start adding your own!
Register with your work email and create a new case study profile for your business.
Related Case Studies.
Case Study
Pepsico's Transformation to Smarter Sales Forecasting with Designer Cloud
PepsiCo, a global consumer packaged goods company, faced a significant challenge in calibrating sales forecasting to supply the right product quantities to its retailers. The sales forecast incorporated a variety of data, including warehouse data, store stock data, and promotional forecast data, all of which were provided by retailers in different file formats and delivered using various methods. The primary challenge was the speed of preparing a sales forecast. With the existing Microsoft Access and Excel-based processes, the time required to prepare this data was so extensive that analysts could only leverage it once a month or not at all. This inefficiency risked under or oversupplying retailers, potentially impacting PepsiCo's business operations and customer relationships.
Case Study
Gexa Energy and AutoGrid's Innovative Demand Response Programs in ERCOT
Gexa Energy, a leading retail electricity provider in Texas, was seeking to introduce new demand response programs for its commercial and industrial customers in the Electric Reliability Council of Texas (ERCOT) market. The challenge was to provide a platform that would allow these customers to lower their energy bills by adjusting their energy consumption during peak energy demand or high wholesale electricity prices. The solution needed to be intelligent, scalable, and offer both manual and automated options for adjusting energy consumption. The demand response programs needed to include Emergency Response Service (ERS), Real-Time Price Response (RTPR), and 4 Coincident Peak (4CP).
Case Study
ZettaNet's Agile Juniper Network Meets Booming Digital Demand in Australia
ZettaNet, a privately-held company based in Perth, Australia, was facing a significant challenge due to the exponential demand for enterprise network, data center, and cloud services in the region. The company's business growth necessitated an upgrade of their core network to meet the increasing bandwidth requirements of their customers. The customers, which primarily include managed service providers, were demanding 1 Gbps connectivity between locations. These service providers then deliver network, data center, cloud, and voice services to a diverse range of customers including local businesses, schools, hospitals, residential communities, and government offices in Western and Southern Australia. The challenge for ZettaNet was to meet this high-capacity network services demand while maintaining profitability.
Case Study
Blue Bottle Coffee Enhances Ordering Accuracy and Reduces Waste with ML-Driven Demand Forecasting
Blue Bottle Coffee (BBC), a global coffee roaster and retailer, faced a significant challenge in managing the supply of pastries across its international network of cafes. The company was using a manual ordering system, where cafe leaders estimated the required quantity of pastries based on historical sales data, current inventory, and growth projections. This system was effective when BBC had a few cafes, but with over 70 cafes worldwide, it became inefficient and inaccurate. The inaccuracies led to either under-ordering, causing sell-outs and customer dissatisfaction, or over-ordering, resulting in food waste and profit loss. The suboptimal utilization of pastries was also affecting BBC's bottom line. Therefore, BBC needed a scalable, precise, and predictive ordering solution to improve pastry ordering accuracy, reduce food waste, and meet its sustainability goals.
Case Study
MARS Incorporated: Leading a Global Digital Transformation
MARS Incorporated, a multinational manufacturer, faced significant challenges in digitizing and standardizing processes across its mid-markets globally. The company's reliance on legacy solutions necessitated continuous, time-consuming upgrades and made data compilation and comparison across different regions and business units difficult. Excel spreadsheets were extensively used for forecasting in smaller markets, leading to disconnected processes, siloed working environments, and increased risk of data inconsistencies and inaccuracies in demand forecasting. MARS also struggled with a lack of visibility across its midmarket footprint, scattered critical data across various systems and spreadsheets, a low degree of automation, and insufficient statistical analytics for demand planning. The absence of a standard process for demand planning made it challenging to consolidate KPIs and gain a comprehensive view of demand trends and supply chain performance.
Case Study
Enterprise AI for Demand Forecasting and Production Scheduling in Global Agribusiness
A global agribusiness and food manufacturer, producing over 80 million pounds of food products per year across eight production lines, faced significant challenges in demand forecasting and production scheduling. The company's only customer, a global retailer, exhibited highly variable demand, leading to discrepancies with the manufacturer's weekly demand forecast. Traditional demand forecasting solutions, based on statistical algorithms, were unable to cope with the short lead times and daily sales orders due to the short shelf life of food products. This resulted in unfulfilled customer orders. Additional rule-based solutions procured to improve production scheduling also failed to optimize schedules and significantly improve manufacturing operations.
