Bristol & West Mortgages is focused on the sale of a range of mortgage products via direct sales, branch network and intermediary channels. Owned by Bank of Ireland Group, Bristol & West Mortgages is one of the biggest lending operations within the Bank's UK Financial Services portfolio. UK Financial Services contributes a significant proportion of the business's profit and totalled over 30% of total Group profit for 2002/03.
Retaining customers at the end of a promotional rate or when their Early Repayment Charge expires is a key issue for most lenders. The customer is free to search and move on to a better mortgage deal elsewhere. Understanding this behaviour and if possible building a model that would predict how customers would behave would give Bristol & West Mortgages an opportunity to develop marketing strategies to reduce customer churn.
In order to address this issue, Bristol & West Mortgages built and tested an SPSS solution based around SPSS' Predictive Analytics software.
“The SPSS Predictive Analytics solution has enabled us to more accurately predict likely customer behaviour and to develop our marketing strategies to reduce the risk of customer churn,” said Stuart Plumley, Customer Analyst at Bristol & West Mortgages.
In order to successfully implement the new solution, Bristol & West Mortgages undertook a Data Mining Proof of Concept exercise with SPSS Professional Services to identify the main requirements, data sources and derivations required to predict likely behaviour. CRISP-DM (Cross Industry Process for Data Mining) was adopted to ensure that the expertise of both parties was transferred into actionable results for the business.
Decision tree predictive models were built to analyse the behaviour of a control group of 10,000 customers. This identified key characteristics that were indicators of likely behaviour to test the performance of the model before deployment.
“Customers were scored with a predicted likelihood of redeeming their mortgage, asking Bristol & West Mortgages to transfer them to a better rate or taking no action,” said Plumley.
Bristol & West Mortgages has started to modify its marketing strategy in accordance with the model's predicted behaviour. Plumley goes on to explain: “Evaluating the results several months after this model was tested has enabled us to improve our understanding of customer behaviour at key points in their mortgage lifecycle. This in turn has enabled us to build retention strategies for our existing customers based on actual behaviour rather than just assumptions on what might happen.
“In the competitive environment in which we operate, having the ability to reduce our levels of customer churn makes a significant difference to our business and so we are very satisfied with the results.”
The complete list of global SPSS success stories can be found here
“Evaluating the results
several months after this
model was tested has
enabled us to improve
our understanding of
customer behaviour at
key points in their
mortgage lifecycle. This
in turn has enabled us
to build retention
strategies for our
existing customers
based on actual
behaviour rather than
just assumptions on
what might happen.”
- Stuart Plumley, Customer
Analyst, Bristol & West
Mortgages.
Key Products and Services
Resources
Demos and Downloads
Was this information useful?