Barclays banks on master data management

Head of data quality and profiling Saul Judah explains how the bank went about implementing MDM

Banking giant Barclays credits its new master data management (MDM) initiative with a long list of business benefits.

These include cutting costs, reducing complaints, reduced risk of fines, better customer data, increased staff productivity, better tackling of fraud and improved financial reporting.

MDM is a method of understanding corporate data and its use to better meet the needs of the business.

Saul Judah, head of data quality and profiling at Barclays, explained exactly how the bank set about putting procedures in place to produce such results.

Effective data management is particularly important in the banking industry for regulatory compliance – reducing the risks of fines – and also for more effectively targeting customer communications.

If customer data is poor, the likelihood of the bank's marketing messages hitting home is reduced.

Due to the size of the company, it's important that employee data is managed well.

"We are working to increase our flexibility in terms of how we are able to use different communication methods to interact with customers. Some prefer to be contacted by phone, some prefer email, some do online banking. We need our data to be of high quality in order to be flexible in the way we interact with customers," said Judah.

Judah also said that it is important to control customer complaint data, which he calls "gold dust".

He said that data that comes through customer complaints needs to recorded accurately so that the organisation can act on the feedback.

Barclays' approach to MDM was to continuously implement a system of defining the data that was relevant to the project, measuring it, analysing it, improving it and putting controls in place.

The first step of the project was to define the business processes that use the key data.

Once that had been defined, Barclays set out to measure the business requirements of the data and current business behaviours – that is, how the data is used.

As part of this process, Judah said that it was also necessary to measure the data quality that is used in key processes.

Once this stage was complete, the MDM team analysed the root causes of why particular sets of data were lacking in quality.

The technologies used to support the project were Cognos at the front end, business information software from Ab Initio and Teradata, and data quality software from Trillium.

A dashboard was created and defined by first defining business process within the organisation.

In order to test the quality of data within the organisation, the MDM team ran queries using the software.

The queries were defined by inputting the business processes and relating them to field names in the database to test the quality of data.

The queries were only run against the most important sources of data. The information on the quality of data was then reported through the dashboard.

And what insights can Judah offer other companies looking to undertake similar projects?

"It's important for staff to understand that this isn't something that can be sorted quickly. It takes time," said Judah.

"You need to ruthlessly come back to the 'So what?' question," he added.

"Always focus on the key drivers: how does what we do affect cost reduction, revenue growth, customer experience, business value and regulatory risk and compliance? And the nature of this work means that a culture of continuous improvement is needed."