HCF provides private health insurance cover for hospital treatment, as well as a wide range of ancillary health and related services including dental, eyecare and ambulance transport services. It has approximately 425,000 members and is one of Australia’s largest registered private health funds.
HCF offers health insurance policies in several tiers – in general, the higher the premium, the more benefits that are applicable. Basic level cover, for example, includes hospital cover with ambulance. Optical, dental, no-gap surgery and other extras cover are all available for additional premiums in different packaged offerings.
Direct mail campaigns are typically used by HCF to up-sell basic members onto higher packages and to cross-sell them other applicable insurance types (for example, life, disability and income protection). However, these campaigns are also costly -- up to $1.50 per mail item in volumes of between 10,000 and 400,000 units per mailout per quarter.
To reduce mail costs, HCF wanted to accurately identify only those members most likely to respond to direct mail campaigns – and target them accordingly.
As Patrick Shearman, General Manager, Information Management at HCF, explains: “In certain scenarios, we may have only a three month window to target people who have put in a claim but subsequently found they weren’t as comprehensively covered for the procedure as they could be. Then they will be more open to respond positively to a targeted offer. We wanted to be able to identify such claim patterns so we could run smaller but more event-driven marketing campaigns.”
Predictive analytics was not only seen by HCF executive management as a way to reduce marketing campaign costs, but also to support additional projects in the business, such as inappropriate practice and claims abuse. “Our executive agreed that predictive analytics would provide the organisation with a preventive measure to limit our exposure to inappropriate claims,” said Shearman.
After evaluating several alternatives, including SAS and Salford Systems, HCF chose SPSS data mining and predictive analytics software, based on past successes with the software. HCF started by using SPSS for optimising membership upgrade campaigns. “SPSS software suited our needs - it had already proven its potential to drive similar marketing campaigns, so there was no reason for us to change. We just didn’t see a point in moving away from it,” said Shearman.
In order to realise the full potential of the SPSS solution, several key internal business area managers were asked to develop models, validate rules and analyse the resulting patterns and trends. In addition, up to 15 analysts attended specialised SPSS training, including four in the Information Management team for centralised support purposes.
“The business managers bring a lot of the subject matter expertise and knowledge required to validate and give context to the rules and results within the overall business framework - discarding patterns they are aware of while mapping newly discovered ones with their own experiences,” said Shearman.
The initial use of the SPSS solution for optimising direct mail campaigns drew “promising” results, including potential savings of $425,000 per year just from being able to predict and identify those members most likely to respond favourably to a targeted campaign. “We found we were able to achieve the same response rates by mailing just 60 per cent of the previous unit volumes – effectively cutting the cost of each mail campaign by approximately 25 per cent,” said Shearman.
Based on these results, HCF will now execute several up- and cross-selling campaigns to carefully selected members. At the same time, HCF is also using SPSS predictive analytics software for a membership retention project. With current reported retention rates of 97 per cent, HCF is using SPSS to determine the predictors for churn and build marketing campaigns to retain members based on the results. “We want to build our retention rates to 98 per cent, which will also increase our recurring revenues,” added Shearman.
We found we were able to achieve the same response rates by mailing just 60 per cent of the previous unit volumes - effectively cutting the cost of each mail campaign by approximately 25 percent cent.
Patrick Shearman,
General Manager
Information Management
HCF
In addition, the health insurer has implemented the SPSS solution to expand the benefits of data mining and predictive analytics into other areas of the business, such as inappropriate practice and benefits management.
“In the first four months, we’ve retrieved in the order of $645,000 from inappropriate practice, which could be as simple as a provider making an error due to the complexity of the claims process. The instances in themselves are small, but they all add up to a significant amount,” concluded Shearman.
Interested in HCF? Download the PDF
Predictive Analytics
can make your organization
more
successful
Contact Us
Key Products and Services
Resources
Demos and Downloads