What going to the dentist taught me about data quality

Photo via Flickr (Creative Commons License) by: Paul Lowry

Dear kind readers, while some of you are reading this very blog post, I will be getting ruthlessly tortured by my maniacal dentist.

Well okay, the truth is that I will simply be getting two cavities filled at my dentist’s office on Thursday morning.  Dr. Blass and her entire staff is far from maniacal—they are, in fact, all very wonderful people. 

I am simply deathly afraid of the object of terror pictured above—the dental drill.  I would argue that this evil object produces one of the most horrifying sounds ever heard in the entire history of humankind.

What does any of this have to do with data quality?

In previous blog posts, I have used a variety of metaphors to compare and contrast the proactive (i.e., defect prevention) and reactive (i.e., data cleansing) approaches to data quality.  With this blog post, I will add an oral hygiene metaphor. 

Brushing and flossing our teeth is defect prevention, where instead of preventing data quality issues before they happen, we are trying to prevent tooth decay and gum disease.  If we neglect these preventative measures (e.g., if, like me, you only floss when you get something stuck in your teeth), then we could develop cavities and gingivitis. 

Removing the decayed portion of a tooth and filling the cavity is data cleansing, where instead of correcting data quality issues after they happen, we are trying to correct the problem before it gets worse (e.g., leads to partial or complete tooth loss). 

Just as data cleansing doesn’t address the root cause (no pun intended) of data quality issues, correcting tooth decay doesn’t address the lapse in oral hygiene that caused it.  However, once the damage is done, corrective action is necessary, or at least preferred before the problem worsens.  Just like data cleansing is often viewed as a considerable cost with little to no ROI, so is getting a cavity filled (especially when, like me, you do not currently have any dental insurance).

I know that this particular metaphor doesn’t really add anything new to what is one of the most deeply polarizing topics for the data quality profession.  However, it is perhaps a more tangible metaphor. 

The vast majority of people have a tendency to neglect their oral hygiene until an obvious (and usually quite physically painful) problem presents itself (e.g., wow, my tooth really hurts, I have to go see the dentist). 

The vast majority of organizations have a tendency to neglect data quality until an obvious (and usually quite financially painful) problem presents itself (e.g., a customer service nightmare, a regulatory compliance failure, or a financial reporting scandal).

My point is that we should all be brushing and flossing our data at least twice a day, and we should all be getting a routine data checkup at least once every six months.  In other words, implement defect prevention whenever and wherever possible, and perform a data quality assessment on a regular basis.

After all, your data probably dislikes data cleansing tools just as much as I dislike dental drills.  Well, almost as much.

Related Posts

Microwavable Data Quality

A Tale of Two Q’s

Hyperactive Data Quality (Second Edition)

The General Theory of Data Quality

Comments

[Garnie Bolling]

OK, I hear the drill, and I know the "drill."

The same old question put ever so carefully:

"Do you want the small amount of discomfort now, or would you like to have a lot of pain later?"

Prevention does go a long way, but sealing a hole in a sinking ship is a bit late.... I like to use the best practice, start small, make a big win now, and yes it's a little discomfort now, but imagine the mental weight that is lifted when you are able to say it was a success (or that filling a cavity was a success vs. a root canal gone wrong).

Oh boy, now I am feeling guilty and need to call my dentist ... like going to confession ... I know I need better data quality, but it did not hurt that bad ...

(Now, can you help me get this high pitch drilling noise out of my head ?)

Thanks again Jim...

[Nicole Carriere (carrni)]

Jim,

Another great blog.

In fact, just like oral hygiene, data quality starts with training and requires a timely routine to ensure efficiency.

Nicole

[Crysta Anderson]

Great post.

So would a root canal be analogous to a "rip and replace" strategy? I shudder at the thought.

[Jim Harris]

Thanks Garnie, Nicole, and Crysta for your comments!

@Garnie — Maybe a software vendor should incorporate the high pitch drilling noise as a sound effect when poor data quality is encountered and data cleansing becomes necessary? The sales pitch could be: "If you don't ever want to hear this sound again, then take better care of your data!" :-)

@Nicole — Excellent point about training. My dentist had to give me a refresher course on proper brushing and flossing technique. Apparently just sliding the floss up and down isn't sufficient, you have rub down into the gum line, and bleeding is an indication that too much bacteria and debris has built up because healthy gums shouldn't bleed from flossing. Likewise with data quality, just buying a tool doesn't do you any good unless you know how to use it, both effectively and efficiently.

@Crysta — Yes, I would have to agree (and also shudder at the thought) that a root canal is analogous to a "rip and replace" strategy, which too many organizations use as an attempt to "solve" data quality by simply buying new technology. It is still the same data and you still need to take care of it.