To paraphrase a former US Supreme Court Justice, “good data quality is sometimes hard to define, but we all know it when we see it”.  It’s even easier to see the effects of poor data quality on our business processes, our sanity, and our stomach lining.  Without an enterprise focus on data quality, data quality issues tend to first become visible deep in the information structure, far from the data’s source and far from a point where a fix can be achieved quickly and efficiently.

When data quality problems are discovered too late, we have few choices, and none of them are ideal.  We can let it go, leaving the problem unfixed for a later effort.  We can go into fire fighting mode, cobbling together a fix that is neither optimal in result nor reusable in ongoing processes.  We can begin the laborious process of tracing the error to its source, diagnosing the problem, and implementing procedures to correct and monitor the root cause for the future.

While the latter approach is preferable, ideally we want to initiate that kind of process before the data quality problem propagates.  Experience has shown that early detection of data quality issues, early development of rules and governance procedures, and consistent monitoring of data quality through dashboards and other simple reports provides great returns to an organization.

So if earlier is better, then how do we push the data quality effort to as early a point as is practical?  The key is getting the business involved.  The business knows the data best.  They live it daily.  The challenge has always been finding an effective way to translate the business rules from business terms to technical terms.  As more hops occur between the business and the ultimate implementation, from legal pad to email to spreadsheet to requirements document to code, the process becomes more costly, more time consuming, and more likely to lose something in the translation.

One of the key goals of Informatica 9, launching in November, is to facilitate business and IT collaboration in a transformative way, with business-initiated rules translated directly into data quality code without any intermediate translations.  I encourage you to participate in the Informatica 9 launch.  I think you’ll like what you see.