Bad Data, Bad Decisions

| Print |

Wednesday, 23 February 2011 15:12

Part 1

There are few adages in the world of computing that are older or more valuable than “Garbage in, garbage out”. The term was coined by an IBM technical instructor circa 1956 and the saying is as relevant today as it was then. The term was used to illustrate that a computer will, without question, process the most nonsensical data and proudly produce nonsensical output.

Similarly, armed with faulty information, human beings are capable of coming to faulty conclusions. And businesses, both large and small, that rely heavily on databases for their information, can make bad decisions when those data are bad.

Imagine the online retailer who restocks inventory based on current shipments of over 1,000 - 50 piece cases or ‘units’ of floor tile only to find out that, when ordering, customers were interpreting ‘unit’ to mean a single tile.

This is an extreme example, but according to author and information expert Larry English, bad data can cost a company anywhere from 10 to 25% of their total revenue.

We’ve established that bad data can lead to bad decisions, so, how do we eliminate bad data? As with anything it is best to eliminate the problem at its source.

So what are the sources of bad data? A data quality survey conducted by The Data Warehousing Institute summarizes them below.

With the major sources of bad data identified, my next article will deal with methods of prevention and correction.

~Steve Barnes

Related articles:

Data Quality Issues Leave Everyone Holding the Bag
Interview with Larry English, Creator of TIQM