We worked closely with the UK Personal Lines marketing group at Zurich Insurance to cross-sell and up-sell to their existing customers, in particular via new channels such as email and text. This involved cross referencing and segmenting customer data to identify customers with appropriate contact details, as well as profiling the types of customers likely to respond to cross selling campaigns. We were responsible for running some of the monthly processes such as the renewal retention model – a SAS predictive model developed to score home and motor insurance customers just ahead of their renewal date, to prioritise which customers get an additional reminder phone call.


At Orange, we used SAS 9 (Base, SQL and Macro) to combine customer data from several different sources to identify customers who had Orange mobile phones and Wanadoo internet accounts. This project formed the first stage of this process, taking customer data from each system, working with a team using Trillium to match addresses. We then used the matched data to produce a series of reports about the shared customers to compare their KPIs against the customer base, to identify any trends in usage or revenue for those customers and then produced a series of statistics and charts to be used by the convergent marketing team in their campaign management.


At Lloyds Banking Group we help integrate the heritage mortgage brands into a consistent data warehouse while also supporting over 80 users in their BAU work, which included mortgage portfolio analysis. As part of the integration project, a new mortgage sales platform was launched and we were involved with the testing and implementation of the system, which has now been rolled out across HBOS as well as Lloyds branches. We then worked with the migration team to bring the credit risk data for 1.8 million HBOS mortgages into CRDM so that the existing Basel II Capital models could be applied across all brands, then uploaded into the new leMans GRID environment.

At Nationwide, we built a framework system using SAS AF for running credit risk models as part of their BASEL II compliance system. Many of the models had been migrated from Excel and still relied upon data to be imported from other systems. By incorporating a number of data-validation steps within the framework, we were able to help the analysts to avoid running models with incomplete or incorrect data.

Central Government

We helped to develop a data mining solution for a government department wishing to combine information from many distinct sources to investigate patterns and links. Our role involved using SAS to extract, transform and load (ETL) the data, combining different datasets for analysis using SAS macros before exporting to i2 Analyst’s Notebook for further analysis and presentation. We also integrated the existing XML data to allow it to be merged with open source data to add value and find links.


For BVG Associates, a small renewable energy consultancy based in Wiltshire, we developed a cost model for use by the Energy Technologies institute to calculate the projected future cost of energy and carbon footprint for new technology developments applying for funding.