When I speak to our members on Data Quality Pro, a lot of their fears revolve around budgetary issues:
- “Will I be able to create a compelling business case for the finance steering committee?”
- “Will our funding run out before we complete phase 1?”
- “How can we hire new staff before we’ve demonstrated value to the business?”
Data quality management is often seen as a cost-base for an organisation but it doesn’t need to be that way. There are many tactics that can be deployed to help improve the quality of your data without calling for a cast of thousands and an executive begging bowl.
I've listed some of my favourite low cost tactics below but feel free to add your own.
Tactic #1 : Make it Easier for People to Register Data Defects at Source
Many data quality issues go unnoticed simply because workers at the coalface have already manually fixed the problem. For example, when my son had his national healthcare identifier duplicated with another child an entire team had to be allocated to fix the problem. If, in a bizarre twist of fate, I had been asked to perform a data quality audit of duplicate data within the health service a few weeks later that issue would never have been found because of the earlier manual intervention.
So how do you track these issues to get a clearer view of where costly cleanup work is taking place?
A simple way I've used in the past is to just offer a dedicated email address internally that people can send information about the fault before they work on it. You can then gather up the incidents each month and track those data quality issues that so often remain hidden from our assessments.
Another option is to mine data from other support teams who may get involved in rectifying data defects that your data quality team can create longer term resolutions for.
Tactic #2 : Archive Your Data Regularly (or at Least Partition from Your Analysis)
This is one of those facepalm mistakes that organisations make and when resolved can dramatically improve your data quality levels or at least give you a far more accurate picture of data quality on the ground.
In one engineering firm we discovered that nearly 20 percent of their plant infrastructure was decommissioned but still being included in data quality reports. This volume of data can seriously skew the accuracy of your data quality findings.
Whether you perform a physical archive in the form of a deletion or simply a virtual archive by adding a termination flag within your data quality processing, make sure you’re not including data that has no relevance to long-term data quality management.
Tactic #3 : Train the Knowledge Workforce in Data Quality Essentials
This is perhaps the most profound tactic I've employed in the last 20 years of working in data quality. There is often an assumption that you have limited resources and finances available to you but by training the very workers who rely on good quality data in basic data quality skills they soon become a vital asset and help you to grow a much needed data quality culture across the organisation.
Having trained junior, part-time data entry staff and transformed them into a crack team of data quality analysts, I can assure you that there are hidden data quality evangelists hiding in every corner of your organisation waiting to be unleashed.
The three tips above are just some of my favourite low cost data quality improvement tactics but here are a few more based on earlier articles on the Data Roundtable:
- How to grow a data quality culture that takes action
- How to improve your data quality history taking
- Re-thinking the design choices of application data quality
- Video tutorial: 5 ways to instantly improve your data profiling performance
- 5 common data quality project mistakes (and how to resolve them)
- Struggling to get started with data quality? Start with data lineage
Why not share some of your favourite low cost improvement tactics? Welcome your ideas below.