It's the data, stupid

Bob Warfield has a great post about how data errors can often be more devastating than code errors -  in this case about AmazonFail.

Much as companies invest in Master Data Management, ETL tools, databases, replication etc, data quality is still inconsistent. Wildly so.

I like the approach some grocery stores take - if the price for a SKU is wrongly coded, they give you the item free or they give you a discount. That's one way to get an intense focus on data quality.

I was talking to someone recently about the challenges of building a global employee master list. Names of villages, father names as middle names, nicknames - technology can only go so far to standardize local nuances. It is actually surprising we do not have more "false positives" in the TSA "no fly" list. On the other hand, the process of unwinding bad data there is pretty cumbersome, and often masked behind reasons of national security.

And then there is the phone number databases that yellow book and other directory services charge for when you call 411..amazing how many numbers are wrong and yet the customer bears the cost because the process of reversing the charges is so cumbersome. 

Not sure you can expect an intense focus on data quality unless there is something punitive about not been intense about data quality.