Data quality key to business intelligence

Business intelligence (BI)  is emerging as a key activity for companies coming out of the economic downturn and a recent paper containing a series of articles published by Computerworld looks at the importance of data quality to BI initiatives. It says the most valuable information that the IT department possesses is the structured data contained within systems such as ERP and CRM, and huge volume of unstructured data collected from emails, call centers, Web sites and collaborative tools.

Unfortunately the quality of much of this data is poor, and the paper suggests that this could cost an enterprise from 15% to 25% of its operating budget. For example, a manufacturing firm that has two identical parts listed with metric and imperial measurements and different parts numbers, or a marketing campaign that sends out hundreds of duplicate flyers.

To take advantage of this data, enterprises need to first focus on data integration and data quality when they look to business intelligence. This will allow the them to take advantage of the accurate information when making decisions, such as lowering marketing campaign costs and increasing revenue.

The paper suggests four tips to improve enterprise data quality:

1. Understand the business context: As ever with IT projects, data quality should be driven by business priorities, so that the IT department can understand the context in which it is required;
2. Discover and profile data: Identify whether you have the right data and assess the state it is in;
3. Monitor data quality: Data quality is for life, don't just carry out a project and let old errors creep back in. Monitor data input and even consider cleaning it on the fly;
4. Implement a data quality methodology: It's key to have a policy and methodology that looks at the overall organizational data quality strategy and business rules.

We're putting together a business intelligence article for the next issue of Real Times, so look out for that.