• United States

Data quality for dummies, Part 2

Dec 13, 20053 mins
Data Center

* How to go about managing data quality

Last week we talked about the challenge of deploying effective data-quality management. The stakes are high. Companies that fail to invest adequately in DQM can expect their expensive information initiatives – like CRM, supply-chain management, ERP, and business analytics – to fail.

As we discussed, the overall strategy for ensuring effective DQM starts with making sure data is accurate when it’s entered, is cleansed, is associated with effective meta-data, and is monitored on an ongoing basis. Sounds great – but how should companies go about doing it?

As with any management project, there are three critical components to a successful DQM initiative: people, process and technology.

Companies need to define a core group of individuals who are responsible for ensuring DQM, and who are empowered to make process and technology changes to do so. The first half of that equation (determining who’s responsible) is often a lot easier than the second (empowering them to make the changes). That’s because it’s much easier to “tag” a group of folks to get yelled at when things go wrong than it is to give them budget and authority to make things go right. Nonetheless, companies need to recognize that they’ll have to be prepared to do both. And keep in mind that the DQM “swat team” can’t work in a vacuum; the group needs to engage a wide range of other individuals, including data owners (typically the lines of business) as well as folks within the finance, compliance and legal organizations.

Companies also need to candidly assess their current DQM processes. Be warned – it’s almost never pretty. Poor DQM processes are the “dirty little secret” of some of the largest and most sophisticated companies in the world. (We won’t discuss the CIO who oversees one of the largest IT budgets in the world and confided that DQM was the single issue that keeps him awake at night.) This means doing three basic things: understanding how data is entered, cleansed, associated with meta-data, and monitored today; deciding how that should happen in the future; and performing a “gap analysis” to pinpoint the areas requiring remediation and change. It’s also important to set in place processes for communication with the groups noted above. As simple as it sounds, holding regular, well-structured meetings to address DQM issues can go a long way to keeping the focus on DQM.

Finally, plan to invest in technology. Nemertes has found that companies that invest in DQM technology are twice as likely to say their DQM efforts are “very” or “extremely” successful – yet more than 70% say they weren’t using such technology. Vendors such as Acxiom, IBM, SAS, FirstLogic, Group1, Information Builders, Innovative Systems, Similarity Systems, and Trillium all have tools and product suites that can help.

The lesson is clear: Invest in DQM – or watch your corporate information initiatives founder.