1 Review the ten commandments for handling information management. For each commandment devise a single sentence which would explain to a business manager why information management is important. Dealing with data is a time-consuming exercise. The need to respond to spam and viruses, or to comply with legislation, can give the impression that data is the enemy.

So turning data into extractable value to reduce time and costs requires a new way of thinking and working.

Companies that do not consider data quality are wasting time and huge amounts of money. Computing brought together a panel of experts to offer 10 tips for making a data quality project work.

The panel:

■ Christine Craven, head of retail information at Abbey (formerly Abbey National bank) ➔

184 CHAPTER 04 INFORMATION MANAGEMENT STRATEGY

■ Tony Rodriguez, chief executive of Avellino software

■ Tom Scampion of Trillium Software

■ Nigel Turner, manager of data quality consulting at BT Exact.

1 Prepare for cultural change Data management is not a quick-fix technology issue. It is a process; a way of thinking that will be a big challenge to many organisations.

It’s a significant lifestyle change, warns Christine Craven. Abbey recently saved £15.2m using a data profiling and analysis product, Avellino Discovery, in conjunction with its datawarehouse.

‘The response of the IT department to businesscritical issues was to offer solutions that were very local, and would have been fine short-term, but would not have satisfied the needs of the business’, she says.

‘We then put the systems areas together with some high-up business people, and came up with a totally different solution which not only satisfied the business-critical issues, but moved information into a central store that was then available to the rest of the organisation.

‘That was a mindset change for IT because they were losing control.’

2 Define what you mean by data quality Concentration on the quality of data is not an exact science. Quality is defined only by business requirements.

‘No company can get 100 per cent control over data quality. That’s an impossible aim’, says Nigel Turner.

‘We don’t own all of it and we all have lots of legacies in our companies and more and more partners.

Data quality is never an end-in-itself goal.’

The starting point for action has to be working out how much valuable data you hold and how you intend to use it, he suggests.

‘If you built a library but forgot to create an index, people would consider that pretty crazy.’

3 Think strategy, not tactics Our thinking about data tends to be based on tactical responses to problems. But value can only be derived from a more strategic approach, says the panel. That means understanding what you have.

‘The problem is that we don’t look at data management in the same way we capacity plan our hardware, network infrastructure or anything else we do in IT’, says Tony Rodriguez.

‘If you can manage your data using the right technologies, you can build a picture. You can find areas to start an initiative where you weren’t even thinking of one.’

4 Ensure clear leadership and ownership Craven is convinced that the key to Abbey’s success is leadership.

‘We have top-down support: the owner of information is a businessman, the chief operating officer, who is the number two in the company’, she says.

The muscle of the highest levels of management ensures that data that does not conform to the rules is outlawed, she adds.

5 Do not allow silo thinking. Centralise policies Most companies have little or no central vision for data management. But a centralised policy is crucial, according to the panel.

‘One of the first things we did at BT was to decide on a best-of-breed toolset’, says Turner.

‘We then gathered together all the people in the company with knowledge and experience of that toolset into one unit. IT departments could not do their own thing and invent their own local solutions.’