Customer Data Integration: Reaching a Single Version of the Truth

BAD DATA’S HIGH COST

At her retirement party in 1983, Grace Hopper predicted that information itself would one day be carried on the books of companies, having become more valuable than the machinery that processed it. The time has come to manage data as a corporate asset. Managers must agree on the value of information to their businesses, and data quality is a key lever in realizing this value.

Answering the question “What does dirty data cost us?” is easier said than done. A more realistic test is to begin with the end in mind, addressing instead critical business issues that may be decomposed into the root cause of bad data.

At a simple level, “What is a lost customer worth• to us?” can be reduced to “Why are we losing customers?” and again to “What percentage of customers are frustrated with billing mistakes?” to “How many billing mistakes do we make in a given month?” to finally, “What are the mistakes and where do they originate?” This last question can magnify specific data problems and their sources.

In a study of 413 manufacturers who shared data with their retailing customers,² there were 2784 data errors that included bad data about products, quantities, and brands. Such data mistakes not only represent a barrier to efficiencies—in this case 44 suppliers were cited as having data so poor that it could sabotage the supply chain—it can result in millions of lost dollars in revenue due to poor merchandising decisions. And with the razor-thin margins already plaguing ...