Business Impact of Data Quality

by Danette McGilvray

July 2014

“What is the point of data quality? How does my company benefit?”
“I just don’t see what data quality means in terms of my business.”

Your organization’s leadership team is dissatisfied because they don’t understand the value of your data quality work. Clearly this is an issue. You need to show how data quality management is necessary and affects your firm in concrete ways.
Decision makers want to know, “What impact does data quality have on our business?” and “Why does it matter?” Another way of saying this is, “What is the value of having information quality?” These are important questions. After all, no one has the money to spend on something that is not worthwhile.
My methodology, called Ten Steps to Quality Data and Trusted Information™, includes business impact techniques that help answer those questions. Using these techniques helps establish the importance and relevance of data quality so that management understands what their money is going to do for them. Using the techniques provides input to making informed decisions about investing in data quality work.
The business impact techniques include both qualitative and quantitative methods that make clear the value of data quality. The techniques are placed on a continuum of relative time and effort necessary to assess business impact – starting with the least complex and less time-consuming technique #1 and moving to the relatively more complex and more time-consuming technique #8. See Table 1 below. One of the reasons the techniques are positioned on this continuum is that too often people think they cannot do anything related to business impact because they don’t have time to fully quantify the cost of low quality data (Technique #7). I wanted to show that you can assess business impact using other techniques that can be equally useful, yet take less time and effort.
Let’s explore a few of these techniques more in depth. Technique 1 - Anecdotes, is one of the simplest, but it can also be one of the most effective. Anecdotes explain concepts of data quality and management by telling a story. Collecting anecdotes and creating stories from those anecdotes is the easiest and most low-cost way of assessing business impact, yet can still produce results. Stories provoke interest so that listeners can relate the impact of data quality to their own experiences. The right story can engage leadership quickly—even without quantitative data. In addition, this technique can be used to clearly explain and provide context for the facts and figures from other techniques.
With Technique 3 - The Five “Whys”, we modify a technique often used for root cause analysis and apply it to get to the real impact of data quality issues. In this process we take a known information quality issue and ask a variation of “Why” five times until we get to the real business impact - growing more specific each time. For example, let’s illustrate this technique against the issue of complaints about poor quality information in reports coming out of the data warehouse.

• Where is the data quality issue showing up?
• In reports coming out of the data warehouse.
• Which reports?
• The weekly sales reports.
• How are the weekly sales reports used?
• Compensation for sales reps is based on these reports.
• Why does that matter?
• If the data are wrong the sales reps will not trust their compensation.Why does that matter?

If the sales reps do not trust their compensation, they spend time creating their own spreadsheets and checking and rechecking the compensation figures—time better spent selling and bringing revenue into the company.
By using the Five “Whys” technique we are able to discuss poor information quality in terms of impact to sales reps and time taken away from revenue-producing activities. This is much more meaningful than saying, “The report is wrong,” or “We have a data quality problem.”
As you can see, the intent is simply to ask the next deeper question. The question might start with “Why” but could also use other interrogatives such as Who, What, Where, When, and How. For example, “Who is impacted by this data quality issue?”, “When and where did the problem occur?” or “How often does this problem arise?” This process should not feel like an interrogation to the person with whom you are speaking, but a collaborative conversation in which you engage and explore the business impact together.
Technique 4 - Benefit vs. Cost Matrix and Technique 5 - Ranking and Prioritization, help to prioritize the needs of the business, and determine which needs should be addressed first. People often ask “Danette, we are talking about business impact techniques, but these are prioritization techniques. Why are they included?” When you prioritize a list of things and something is ranked as #1 and something else is #10, what are you saying? You are saying that whatever is #1 is more important or has a higher impact than whatever is #10. With data quality we are always asked to do more than we have time and resources for. Therefore it is critical to have good techniques to show business impact through prioritization.
Most people think the only way to show business impact is through fully quantifying the impact, which is done using Technique 7 – Cost of Low Quality Data. This technique is what most people think about first when asked to show business impact. Yet all of the techniques mentioned in Table 1 can be successful on their own or in combination with one another.
As mentioned, the techniques vary in both time needed and in relative complexity to assess. Understanding where they fall on the continuum offers the ability to better choose the ones most applicable in any given situation. Use the techniques you think will be most worthwhile based on the amount of time and resources you have available, but know that you can always do something related to business impact. The importance of being able to show business impact cannot be stressed enough. It is essential to getting any kind of support for data and information quality—be it time, resources, money, or expertise.

Note: Portions of this article contain material from the book Executing Data Quality Projects: Ten Steps to Quality Data and Trusted Information™ by Danette McGilvray, published by Morgan Kaufmann Publishers, copyright 2008 Elsevier, Inc.
This article Copyright 2013 by Danette McGilvray, Granite Falls Consulting, Inc. All rights reserved worldwide.