Tuesday, July 6, 2010

Prevention or cure?

I was looking through a pile of my old and dusty university essays a few weeks ago, nicely typed on a 40-year old manual typewriter (at that time the university's only computer was in a huge, well guarded room and the only way we were allowed to interact with it was with punch cards ...) and I found an essay with this title:

"Regional Water Authorities in Britain are Dominated by Engineers Trained Largely to Solve Water Supply Problems by Constructing New Facilities Rather Than by Minimising the Need for Them" (D.J. Parker and E.C. Penning-Rowsell). Discuss."

I had discussed, as directed, and agreed: instead of tempering our profligate and ever-increasing use of water, we just kept tapping into new resources to increase supply.

It wasn't just water that had this issue. And little has changed over the past 30 years.

Looking around, you see this pattern almost everywhere. Health services, for example, spend a little on prevention and a huge amount on curing. Police try to solve crimes after they occur but rarely attempt to prevent the crimes from occurring (and in many countries they are not allowed so to do). In fact, our whole society is based on the use/consume/experience now and the resolve/cure/clean up later paradigm.

So it's hardly surprising that businesses work the same way when it comes to data quality. Like the water authorities, they are dominated by people who are trained (and indoctrinated) to resolve problems as they arise rather than to prevent the problems from arising; and when they see the problems they envisage only technical solutions without any consideration for process or business structure changes. The bigger, more expensive and flashier the product, the more likely it is to be bought, regardless of its effectiveness at reducing the problem.

Shifting spending to prevention will reduce spending on the cure, and we'd be healthier for it - stopping us from taking up smoking will always be better than treating us for lung cancer. There'll always be a need to cures - like our bodies, data decays and we have to work on it to keep it healthy - but prevention works better. And is cheaper.

1 comment:

JohnIMM said...

Excellent post, Graham.

You so echo my sentiments. The data quality world needs to move away from Quality Control (which is all about discovering and removing defects) into the world of Quality Assurance - which is all about creating zero defects!

Data Quality Assurance needs to be built into the Business Functions (and here I mean a Function as an activity, not a department) and Processes that create or amend the data.

The trouble is that Quality Assurance is invisible and has no heroes. Zero defects are created and so there are zero defects to be found.

In the current world, data defects are created all the time and finding them can so exciting and heroic! The better hidden the defects are, the harder it is to find them, the clever the person and the tools have to be and the more rewarding it is when they are uncovered!

Who would want to give up all of this excitement and achievement for the boring world of Quality Assurance?