Nerd Appeal or Boardroom Fare?
Posted by 
Share this Post
Data nerds should be able to talk about benefits that matter to the boardroom
I don't know how many of you are members of the International Association of Information and Data Quality (IAIDQ.org), but I've been a member for a couple of years, as are friends and colleagues.
Recently, one of their key contributors and members, C. Lwanga Yonke, posed an interesting question to the IAIDQ LinkedIn group, to which there are dozens of thoughtful answers.
Lwanga's question: "Why is it so hard for C-level executives to understand the strategic importance of quality data? Or is it that they understand it, but they have little faith that DQ efforts will be successful?
Most executives with whom I've worked over the decades are pretty smart folks who have the ability to understand quite abstract concepts and make significant business decisions based on "gut feel." (How many of you read Malcolm Gladwell's Blink?) They also have to constantly prioritize between the threat of the month and the next big thing.
For successful companies there is more than dumb luck and usually these organizations have executives who know how to make sound business decisions based both on fact and their gut.
So why don't they rush to spend precious resources investing in improving data quality? Don't they get it?
In my opinion, those of us who have labored in IT and earned our stripes in data quality related disciplines have done a poor job in relaying the business value of data quality, strategic or otherwise.
We're quite good at discussing percentages of duplicate records down to three decimal points, but are we able to tie that abstract concept to the gain or loss of revenues, or the impact on profitability? I'd say that no, we have not done a good job of that.
In fact, I'd say we've insisted on answering fairly straightforward questions of "what's in this for me?" with lots of technical jargon.
How many of you are fairly comfortable discussing issues of referential integrity, the percentage of records with a strange gender code that correlates to erroneous data in a third address line, and the number of records that seem to have outlier values compared to the rest of a set?
How many of us get excited discussing the benefits of data profiling, for example? I'd guess that most people I know in the industry would be fairly comfortable discussing these topics.
But how many of you would feel confident in discussing the revenue impacts of poor billing addresses and be able to use real metrics to back up your assertions? How many could quantify the hit to profitability caused by poor identify management? How many would blanch if asked to explain the payback of a data quality tool in terms of revenue, install base, gross margin or market share - using real numbers and not hypotheses?
I've been discussing metrics quite a bit in the context of data governance. I believe that the policies we write to help us govern the quality of our data should be measurable, and I think most people agree.
The question I would like to pose here is this: are you going to define metrics that make the IT data analyst in you giddy, or are you going to strive to define metrics that will be meaningful in the board room?
I'd LOVE to hear your answers!
6 Responses »
Trackbacks
Leave a Response
Entries(RSS)
Great post Marty,
I too am a member of the IAIDQ. I have had the pleasure of attending some of Lwanga's conference presentations. He is a great thought-provoker and ambassador for the international data quality community.
I enjoyed Gladwell book Blink because it really calls into question our general tendency to over analyze things. The book provides many examples when “experts” struggled with finding evidence to support a decision and felt uncomfortable “going with their gut” even though their intuition later proved valid. Of course, the book is well balanced in that it also provides examples of the opposite when gut instincts were proven wrong and decisions made in the moment proved to be impulsive and in some cases dangerously so.
I think where many (both Nerds and Executives) trip up when trying to create effective metrics is that they only seek out the data that supports their perspective--otherwise known as confirmation bias--and often conveniently ignore the importance of either data quality (Executives) or business relevance (Nerds).
Back to your point, I agree that as data quality professionals, we generally do a very poor job in relaying the business value of data quality.
I also agree that most data quality metrics are designed for nerd appeal and not for boardroom fare. However, I have also seen the opposite, where more "executive friendly" and "business relevant" metrics were created but were either oversimplified, hypothetical, or based on very specious "real numbers."
The path of least resistance is the great enemy of all metrics.
The cliche of "if you can't measure it, then you can't manage it" is closely related to "if you can easily measure it, then it probably doesn't mean anything."
The Nerds and the Executives need to do the hard work that is no fun for either side, but it essential for success -- and that is sit down together and determine what metrics are meaningful, do represent tangible business relevance, and can be both effectively measured and efficiently managed.
Sure, it's a lot easier for the Nerds to "geek-out with mathematics" and the Executives to "go with their gut." But who ever said this was supposed to be easy?
Best Regards,
Jim
P.S. Nerds Rule
@Marty
In my opinion, those of us who have labored in IT and earned our stripes in data quality related disciplines have done a poor job in relaying the business value of data quality, strategic or otherwise
Good post.
It's often chicken and egg question, at least from my experience.
If the data is bad, then how can you assess the effects? If you can't assess the business effects of a DQ effort, then why should I bother?
@Jim
The cliche of "if you can't measure it, then you can't manage it" is closely related to "if you can easily measure it, then it probably doesn't mean anything."
Very well put. It's quite the dilemma. Too many people view thinks along the lines of Blink, in my view. One time in a meeting discussing data in relation to employee turnover, a senior person once said, "I don't need data to tell me why people leave."
How do you argue with that mindset?
Great post! Defining metrics that interest executives has always been challenging. Some of the things I have found they are interested in are:
Savings in operational costs
Greater business effi ciency from a more integrated supply chain
Better strategic planning based on more accurate analysis and forecasting
Higher customer satisfaction and retention
The challenge is how to measure these things. To measure savings in things like operational costs and greater business efficiency you could measure how much time is spent on re-active data cleansing activities. This could be relatively simple to measure if resource costs for these activities are tracked. To show greater business efficiency, better strategic planning and higher customer satisfaction you might have to go with before and after surveys. I have found many executives aren't really interested in results measured with surveys.
It would also help if there were more dedicated resources allocated to success measures, benefits realization and the development of business cases. I have found that most often, the measures are an afterthought that little time is devoted to.
Thank you!
Marty, and excellent article, once again. Like Phil and Jim's comment, if it's not easy, then why measure/manage to it.
I have seen recently that tooling is VERY important. Gone are the days of just know what the data looks like, what is more important, is how it is used (or not), how it is created (or not) and how to leverage it (or the opposite).
What is needed is a set of tooling that allows the Nerds to collect and create analytics of the data. Being able to have the right evidence to "Tune In" on the right combination of quickly showing value and balance the amount of effort it is to create the right data (match, merge, organize).
It is refreshing to see that there are leaders in this space of DQ... now sharing this information to the business / board, well, it works.... it works well...
Thanks again.
Jill! Thanks for the great input! Very helpful to suggest a dedicated resources to "own" the ongoing measurements & metrics trends.
Garnie! Great to hear from you again - you dropped off the radar screen for a while... Thanks for your feedback and input!
Cheers, everyone!