Data Quality – Why “close enough” isn’t “good enough”

close enough is not good enough

I was discussing the data quality vendor landscape with a colleague the other day and he drew a very interesting parallel, an analogy about solutions and performance I hadn’t heard before. Frankly, on the surface it was almost comical — but the more I thought about it, the more I realized the significance.

So picture yourself in this situation…

You buy an airline ticket from NY to LA for a wedding on a weekend. You depart on time, but when you land you find that you’ve arrived in Reno Nevada – Not LA! Certainly not your destination. You then come to learn that the plane you boarded was incapable of making the full distance from NY to LA.

Are you satisfied? Are you happy? It’s a silly question with a profoundly obvious “no” answer, because what would be the point?

Here’s the parallel.

Would you invest in a data matching and data quality logic that doesn’t perform well? This question should be as absurd as the former with an equally obvious answer — but it often isn’t.

For some reason, too often technology buyers (especially those in procurement) often don’t take that same view of buying technology that they do in so many other decisions they make in their daily lives. They’re tasked to find a solution that does ‘X’  — and to be fair, let’s face it, vendor solutions often sound a lot a like. Of course the reality is very different.

Now, I could spew off a bunch of statistics and percentages that try to convey how important it is to maximize match rates while minimizing false positives and missed matches, but this isn’t that kind of story.

If you think the story is about the flight, or about quantifying the number of missed matches from one solution to another – you’d be forgetting the real point. You didn’t buy a flight because you needed a ticket, and you didn’t buy it because you wanted to go to LA. You bought the ticket because you wanted to attend your best friends wedding!

Data quality is a function of a bigger picture. Data quality is about protecting and measuring business decisions and insights.

The real story that needs to be heard is, why “every single” record is important. Why skimping on data matching accuracy and failing to reach your destination in the name of budget or the effort of changing vendors is as silly as flying to LA and landing in Reno – and being happy with that!

The link between a name and a person is inextricable to the point a name is a person, but it could…