Strategy clinic: How can I make users supply quality data?

The IT team has been criticised over our customer relationship management software. But the real problem is the quality of the...

The IT team has been criticised over our customer relationship management software. But the real problem is the quality of the data supplied by users. How do we make users understand the importance of data quality, especially that supplying old data created for other purposes is not appropriate?

The business case should clarify the technical case

For a business solution to be successful, the business and the users need to be engaged throughout the whole process. If the CRM project has been exclusively an IT project, this will be the root of the problem.

To successfully install any application, an appropriate business case should exist that clarifies the technical case. The business case will clarify business need, the critical success factors, benefits realisation, dependencies and risks.

You need to establish whether the users understand and have bought into the need for the CRM system. Hold some workshops and market the need. Show the users why the old data will not help them do their jobs.

Make sure you have not fallen in to the trap of enforcing processes that, although great in theory, are unworkable in reality.

Roger Rawlinson NCC Group

Taking the heat out of the political aspects

Tackling data quality is a politically fraught task and can easily put an IT manager in the bad books of senior managers.

Criticism of operational data tends to be perceived as criticism of the staff who produced it. Questioning the quality of management information exposes weaknesses in managerial decision-making processes. Also, data quality has to be improved by changing business processes and working practices, which means hard work for the business.

If you want to tackle data quality head on, be prepared to make yourself unpopular and do not expect any rewards. If data quality is poor, the business will find fault with your poor IT; if data quality is good, it will be because it is their data.

Many data quality problems are no more than mismatches between data quality as understood at the point of data capture and by decision-makers at the new point of data use. This should be seen in a positive light - business users are adapting by trying to use data in new ways.

Engage managers in thinking about the different ways in which the information they have could be improved to gain an appreciation of what "fit for purpose" means for them.

You can then identify the business and IT changes needed to deliver information to the standards they require. Feed your assessment back to decision-makers in terms of their own needs, explaining the costs and business change required.

A few small wins will raise awareness among managers of the need to take a more strategic approach to managing one of their most valuable assets.

Christopher Hemmingway, Cranfield School of Management

Successful change comes from business and IT unity

Many organisations still struggle to bridge the business/IT divide, resulting in frustrated IT departments and dissatisfied users. Successful business change programmes require business and IT teams to understand and fulfil their responsibilities. Critical success factors include:

  • Visible leadership by user management, fully briefed in the programme and its roll-out
  • Early involvement of users
  • Quick wins to reinforce benefits
  • Clear responsibilities and commitment to time and effort required from all parties
  • Comprehensive communication plans
  • Considering bringing "influential cynics" into the team to champion the roll-out.

Changing the hearts and minds of users after a system is live costs 10 times as much as doing so before. It can be done, but lots of good systems have been dropped because users have refused to accept the change.

I suggest you set up a small team of influential people from IT and the business department involved. Come up with a realistic action plan and get the management resources needed to fix it. A CRM system with inaccurate data can be a very expensive business mistake.

Steven Bell, partner, Ernst & Young

Consider data quality issues before, not after, investment

The problem with data quality is that although everyone agrees it is a good thing in principle, it is not simple to achieve in practice. For example, the people who create business data are often not the people that have to use it, and data created years ago may now need to be used in ways the company never envisaged.

Lead your company through the next maturity step - considering the issue of data quality before committing to an investment, not afterwards. Identify what data will be critical to achieving the targeted return on investment and check that the company either has that data at sufficient quality, or can realistically invest in improving and enhancing the existing data.

Chris Potts, director, Dominic Barrow

Time to focus on the 'I' part of the IT director's role

Who is responsible for the quality of data? It is easy to say that the users are, and it then seems very unfair that they should blame the IT department for their failure to supply data that is fit for purpose.

However, the alternative view is that IT is accountable for the success of the whole business project. Ask what the IT department can do to make users understand the importance of the project.

  • Ensure that a business project manager is appointed who understands the importance of data quality.
  • Provide tools to assess the quality of the data, not just during testing but before, during and after the project.
  • Ensure there is effective user acceptance testing that identifies data quality issues.
  • IT can encourage the appointment of data stewards who have an ongoing responsibility for data quality.

You can treat this problem as an opportunity to focus on the "I" part of the IT role. This is a major challenge, and if the "T" part is working, it can be a very positive one.

Sharm Manwani, Henley Management College

Revisit what was agreed in the project specification

Surely one of the principles of CRM systems is that they manage data to better serve the customer and to provide new opportunities for the business.

You and the users need to revisit what was agreed in the project specification, because either you are wrong in that you did not specify the data quality needed, or they are at fault in not providing pre-agreed quality.

It just stresses again the importance of defining requirements before proceeding to application delivery.

Robin Laidlaw, Computer Weekly 500 Club

Ultimate responsibility lies with the business

CRM done badly is worse than not doing it at all. If this is not already understood, you need to communicate it quickly and concisely to users at every level. Identify areas of information accountability and clarify where the ownership and responsibility lies.

I recommend that it lies with the business - the IT department should only be the guardian of the information. Work with data owners to understand what you have and how you need to use it. Create definitions, build a model and set out rules. User acceptance and testing is essential.

Ollie Ross, The Corporate IT Forum

The experts

Computer Weekly has put together a panel of experts. You can draw on their specialist knowledge to solve a problem. E-mail your questions (or your own solution to this question) to computer.weekly@rbi.co.uk

NCC Group www.nccglobal.com

Ernst & Young www.ey.com

Cranfield School of Management www.cranfield.ac.uk/som

Computer Weekly 500 Club www.cw500.co.uk

Henley Management College www.henleymc.ac.uk

British Computer Society www.bcs.org.uk/elite

Impact www.impact-sharing.com

The Corporate ITForum www.tif.co.uk

Dominic Barrow www.dominicbarrow.com

Put your question to the panel
If you have a question you would like answered, e-mail computer.weekly@rbi.co.uk.

Read more on Database software

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close