SUCCESSFUL PROJECTS
Santam is South Africa’s largest short-term insurance company. With more than 650 000 policy holders, and assets of more than R17 billion under management, the company enjoys a market share of more than 22%. Santam operates in a market where fraudulent activity accounts for an estimated 6 to 10% of all premium revenues.
Santam wanted to improve its’ service to customers by settling claims faster, and keeping premiums low. To achieve this, they needed to maximize operational efficiency and find more resourceful ways to combat fraud. Santam worked with OLSPS to design a claims segmentation solution using IBM SPSS Software.
Our solution needed to integrate seamlessly with the current Santam claims process, and allow for real-time scoring of each claim with a fast response rate. Each claim is automatically scored according to its risk level, and then moved to the appropriate processing channel, for either settlement or further investigation. This allows high-risk claims to be sent to the investigation unit, while low-risk claims can be processed quickly, improving the overall customer service.
The solution offered by the OLSPS team:
• Enhanced Santam’s ability to detect fraud, saving them R17 million within four months of implementation
• Improved customer service by enabling legitimate claims to be settled within an hour, which wasmore than 70 times faster than previously
For more information you can download the case study from Nucleus, or watch the following video.
Millicom International Cellular, also known as Tigo, is a mobile network provider that services more than 30 million customers in 13 emerging markets across both Latin America and Africa. Tigo specialises in affordable, widely accessible, and readily available prepaid cellular telephone services.
Tigo approached OLSPS to assist them with two business challenges:
• To understand their customers profile, and how customers used their services
• To predict the customers that are most likely to leave them in favour of competitor services
The first business problem was solved with the OLSPS Customer Segmentation Solution, which identifies different behavioural profiles within the customer base. The information gained from these profiles enables Tigo to conduct highly targeted marketing campaigns. This information is also used to develop products and services to fit each profile.
The second business challenge was solved with the OLSPS Churn Solution. Most of Tigo’s customers make use of prepaid services, which allow them to move to another service providerwithout any warning. This poses a problem for Tigo, as the cost to acquire new customers is considerably higher than retaining existing customers. The OLSPS Churn Solution scores each of the customers to predict the probability of the customers churning before they do.
A similar solution to that of Santam was implemented for Mutual & Federal using SAS software. Working closely with the Mutual & Federal data science team, we managed to produce a fully integrated, real-time solution that flagged fraudulent claims in the motor insurance sector. After implementation of the solution customised training of the Mutual & Federal data science team enabled the client to maintain and manage the solution internally.
P-Cubed, a leading management consulting firm, specialises in delivering critical and complex programmes and portfolios for governments, agencies, and major institutions across all industry sectors. P-Cubed had a requirement for an automated scorecard process to facilitate credit-based response modelling. OLSPS developed a scorecard within IBM SPSS Modeler, which presents the credit risk scoring as functional nodes within the IBM SPSS Modeler environment.
Overall, the OLSPS Scorecard Solution provided for rapid automated development of scorecards alongside assessment utilities, and is used as a key component of P-Cubed’s business process.
Nedbank is one of South Africa’s largest retail banks. With the fourth largest customer base, it is critically important to monitor customer sentiment at all times. With this in mind, Nedbank approached OLSPS, to develop a social media monitoring tool built on IBM SPSS Software. Nedbank uses this tool to collect data from the online interactions between them and their clients. These interactions happen on different social media feeds, including news feeds, blogs, Twitter, Facebook, and other relevant sites, resulting in ‘unstructured’ data.
The OLSPS Sentiment Analysis Solution makes use of text mining to gather insight from the interactions to transform the data into structured data. The data are used to drive customer engagements, and upsell products and services. For more information, see the following video.
Major SA Retailer
We implemented the OLSPS Churn Solution for a large South African-based multibillion dollar retailer which employs close to 32 000 people. They approached OLSPS to help with extracting value from their customer datasets.
The first solution that we developed was a churn model for their cellular client base. This model predicts when a customer is likely to leave their network and helps them to design targeted marketing campaigns aimed at retaining their high value customers. The model was proven to identify the churners with a high level of accuracy.
Metropolitan is one of the largest medical aid providers in South Africa. Medical aids expose themselves to a variety of risks from different sources. One of the most costly risks is that of pharmacies or doctors conspiring to defraud the medical aid. While this fraud can be seen to be relatively small on a claim by claim basis, the volume of claims associated with some fraudulent doctors or pharmacies can be significant and can pose threat to the bottom line of the medical aid.
OLSPS implemented a fraud detection solution for Metropolitan using IBM-SPSS Modeler which focused on fraudulent activity by both doctors and pharmacies. The solution does not operate on a claim by claim basis but rather looks at groups of claims associated with a particular doctor or pharmacy. The outcome of the solution is the identification of the propensity for doctors and pharmacies to submit fraudulent medical aid claims, and facilitates the production of a short list of doctors and pharmacies whose claims should be forensically investigated.
Major Australian Bank
A major Australian bank appointed OLSPS Workforce, a sister company to OLSPS, to analyse the link between mindsets & behaviours of the bank staff, and bottom-line performance. Staff answered short surveys about themselves and their peers over the course of ten weeks, during which OLSPS Workforce accurately measured the inherent, clientspecific behavioural framework within the bank.
OLSPS Workforce then built machine learning algorithm that mapped the inherent mindsets & behaviours to bottom-line performance of the business unit (revenue to plan). This algorithm provided central, data-driven “Ideal behavioural profile”, linked directly to performance, which directed ongoing development from the top down, and the bottom up. Furthermore, the ideal profile significantly improved recruitment efforts. The engagement has since been expanded through to 25 different regions across the client’s operational infrastructure.