In a true story that I recently read in the book Switch: How to Change Things When Change Is Hard by Chip and Dan Heath, back in 2004, Donald Berwick, a doctor and the CEO of the Institute for Healthcare Improvement, had some ideas about how to reduce the defect rate in healthcare, which, unlike the vast majority of data defects, was resulting in unnecessary patient deaths.
One common defect was deaths caused by medication mistakes, such as post-surgical patients failing to receive their antibiotics in the specified time, and another common defect was mismanaging patients on ventilators, resulting in death from pneumonia.
Although Berwick initially laid out a great plan for taking action, which proposed very specific process improvements, and was supported by essentially indisputable research, few changes were actually being implemented. After all, his small, not-for-profit organization had only 75 employees, and had no ability whatsoever to force any changes on the healthcare industry.
So, what did Berwick do? On December 14, 2004, in a speech that he delivered to a room full of hospital administrators at a major healthcare industry conference, he declared:
“Here is what I think we should do. I think we should save 100,000 lives.
And I think we should do that by June 14, 2006—18 months from today.
Some is not a number and soon is not a time.
Here’s the number: 100,000.
Here’s the time: June 14, 2006—9 a.m.”
The crowd was astonished. The goal was daunting. Of course, all the hospital administrators agreed with the goal to save lives, but for a hospital to reduce its defect rate, it has to first acknowledge having a defect rate. In other words, it has to admit that some patients are dying needless deaths. And, of course, the hospital lawyers are not keen to put this admission on the record.
Whenever an organization’s data quality problems are discussed, it is very common to encounter data denial. Most often, this is a natural self-defense mechanism for the people responsible for business processes, technology, and data—and understandable because of the simple fact that nobody likes to be blamed (or feel blamed) for causing or failing to fix the data quality problems.
But data denial can also doom a data quality improvement initiative from the very beginning. Of course, everyone will agree that ensuring high quality data is being used to make critical daily business decisions is vitally important to corporate success, but for an organization to reduce its data defects, it has to first acknowledge having data defects.
In other words, the organization has to admit that some business decisions are mistakes being made based on poor quality data.
In his excellent recent blog post Half Measures, Phil Simon discussed the compromises often made during data quality initiatives, half measures such as “cleaning up some of the data, postponing parts of the data cleanup efforts, and taking a wait and see approach as more issues are unearthed.”
Although, as Phil explained, it is understandable that different individuals and factions within large organizations will have vested interests in taking action, just as others are biased towards maintaining the status quo, “don’t wait for the perfect time to cleanse your data—there isn’t any. Find a good time and do what you can.”
Remarkable Data Quality
As Seth Godin explained in his remarkable book Purple Cow: Transform Your Business by Being Remarkable, the opposite of remarkable is not bad or mediocre or poorly done. The opposite of remarkable is very good.
In other words, you must first accept that your organization has data defects, but most important, since some is not a number and soon is not a time, you must set specific data quality goals and specific times by when you will meet (or exceed) your goals.
So, what happened with Berwick’s goal? Eighteen months later, at the exact moment he’d promised to return—June 14, 2006, at 9 a.m.—Berwick took the stage again at the same major healthcare industry conference, and announced the results:
“Hospitals enrolled in the 100,000 Lives Campaign have collectively prevented an estimated 122,300 avoidable deaths and, as importantly, have begun to institutionalize new standards of care that will continue to save lives and improve health outcomes into the future.”
Although improving your organization’s data quality—unlike reducing defect rates in healthcare—isn’t a matter of life and death, remarkable data quality is becoming a matter of corporate survival in today’s highly competitive and rapidly evolving world.
Perfect data quality is impossible—but remarkable data quality is not. Be remarkable.