Data infrastructure optimization, availability & security software
Data integration & quality software
The Next Wave of technology & innovation

Anatomy of a Data Quality Failure

What does a data quality failure look like in the real world? Americans across the country found out recently when they received marketing offers for a product that was completely misaligned with their ages and genders.

Anatomy of a Data Quality Failure: Incorrect age and gender

What Happened in this Data Quality Failure

We won’t name the company that made this mistake because a mistake like this could happen to any company. The problem arose from a data quality shortcoming. Any business with lots of data can easily face this type of challenge.

Suffice it to say, however, that the company is a major retailer of personal care products. In a nutshell, this is what happened:

  • The company runs a program where it delivers free products to young men on their eighteenth birthdays.
  • Recently, the company delivered several of these free products to people who were not at all its target audience. Some were middle-aged women. Others were middle-aged men. At least one was a woman who had recently turned eighteen, but had no use for a product designed for men.

While this data quality failure didn’t have huge financial consequences, it was an embarrassment for the company. It also amounted to wasted marketing and product resources.

Download our eBook: The New Rules for Your Data Landscape

Data Quality’s Role

While the company has not explained exactly why it sent free materials to the totally wrong recipients, it’s a safe bet that data quality was at the root of the problem.

What likely happened is this: The company presumably uses a database of names, addresses, gender information and birth dates to find young men who are about to turn eighteen and should receive a free product. There were mistakes in that data set that led the company to misidentify a number of recipients.

The problem could have been a mismatch between names and addresses. Perhaps some birth dates or gender details were entered incorrectly.

Data Quality Failure: Incorrect Age and Gender

Or maybe there was an address overlap problem caused by eighteen-year-old men having similar street addresses as other people. For instance, a fifty-year-old woman named Jane Doe who lives at 123 Main Street in Troy, Michigan might inadvertently have received the product that was supposed to go to a John Doe who lives at 123 Main Street in Troy, New York.

When you’re trying to market to every eighteen-year-old man in America mistakes like this are easy to make. Something like 5,500 men turn eighteen each day in the United States, and keeping track of every one of them is hard.

While the exact nature of this data quality mistake is hard for people outside of the company to track down, the solution for avoiding similar problems in the future is clear. Data quality tools that can recognize data mismatches, erroneous entries and other problems within data sets can identify the types of mistakes that likely led to this company’s embarrassing marketing campaign.

To learn more about how you can prevent this kind of data quality failure and others, be sure to check out tomorrow’s follow up post, “Understanding Data Quality: How Data Quality Errors Arise.”

The data workflow is shifting from IT to the business, be sure you know The New Rules for Your Data Landscape!


Leave a Comment

Related Posts