The perception of quality and the reliability of data assets can change with time. It only takes one incident or error for someone to doubt the information on one’s neatly printed payslip (which is almost always correct.) That one mistake can drive someone to verify every subsequent payslip. It is easy to imagine scenarios where a single bad master data record can disturb a well related high quality data asset. Bad perceptions and negative user opinions can drastically reduce the effective usage of even the most reliable and well integrated data assets.
The perception of quality and the reliability of data assets can change with time. It only takes one incident or error for someone to doubt the information on one’s neatly printed payslip (which is almost always correct.) That one mistake can drive someone to verify every subsequent payslip. It is easy to imagine scenarios where a single bad master data record can disturb a well related high quality data asset. Bad perceptions and negative user opinions can drastically reduce the effective usage of even the most reliable and well integrated data assets.
In Gerald M. Weinberg’s article Agile and the Definition of Quality, the author details how quality can be quite relative and abstract. He argues that the definition of quality is more emotionally and politically driven and is more of a relative thing to individuals.
Putting these thoughts into the data quality perspective, the following bullet points stand tall:
- Data quality is to be driven from the end users’ point of view – while standards, rules and infrastructure best practices to store and retrieve data enable good data quality, the focus on the ultimate consumers of data should not be lost.
- Opinions and impressions on data reliability counts – often in organizations, people spread the stories around severity 1 tickets especially if they are related to the reliability of data. History of pains caused by past incidents due to bad data qualityare fresh in the minds of the people affected. Before anyone can trust and reliabilty of the data, the fears or past impressions have to be addressed.
- Insight into active diverse views of data by participating data consumers – data quality initiatives should encompass frequent examination of how the organizational data assets are seen from many diverse eyes of the end users. The program needs to know which data is most critical, frequently accessed, least understood and depended up on to make decisions.
Overall, data as an asset quickly becomes a liability when it is not used by anyone. So, the opinions, impressions and confidence of the consumers on the organizational data is perhaps the single most critical aspect for a good data quality program.
