In his blog post “Data Quality and The Middle Way” on the Data Flux Community of Experts, Jim Harris states ‘Data quality practitioners must learn to strive for continuous improvement, but without losing themselves in an ideal such as data perfection’

I couldn’t agree more, striving for continuous improvement should be the ongoing goal for any data governance initiative.

It is too often that data governance initiatives focus purely on the initial goals and deliverables, and forget about planning for continuous improvement. Once the project manager leaves and the team is disbanded, with no continued monitoring and measurement in place, the initial successes are wiped out as the in-life deployment flounders in a sea of change and static advancement.

Deploying a process and structure of active monitoring and measurement, with defined responsibility and accountability, will ensure continued success for any data governance initiative. Furthermore, by ensuring a culture and framework for continuous improvement the initiatives will grow and continue to deliver over and above initial expectations.

What is Continuous Improvement

There is no set framework for continuous improvement. Traditionally it comes as a product of a business improvement or performance management program. A continuous improvement framework needs to fit your organization and align with your goals and ambitions, it is not something that can be bought off the shelf.

My favorite definition of Continuous Improvement is:

“A structured measurement driven process that continually reviews and improves performance.”

Breaking that definition down, we can identify that we need to deliver something which continually drives business improvement through a structured, measurement driven process. Looking at like this provides a number of main focus points for a continuous improvement framework.

I prefer to focus on three of them, being:

  • Structure
  • Measurement
  • Continuous Review

Structure

The diagram below (from the UK Audit Commission) shows the typical steps in translating strategy performance measures through a business improvement model.

clip_image002

Having a defined approach, like this, to your performance measurement and improvement allows for structure and control, with what I think are the key products namely responsibility and accountability.

The fact that performance management impacts more on your staff than any other resource, them knowing who is responsible and accountable for each point in the data governance process is essential to successful in-life management, control and improvement.

Measurement

In a previous blog post on ‘Data Governance at the Front Line’ I concluded that:

“Data ownership needs to be part of the psyche of every employee, make it part of their incentives, set it as a key KPI in the performance management of the organization”

Which aligns with another phrase I like and often use:

“You can’t change what you don’t measure!”

The key word here is ‘Measure’. Measurement is a vital factor in driving change and improvement in any business process or initiative. However before we set any KPI’s we need to understand what we are measuring, we do this by translating business strategy into performance measures and targets.

The dimensions you measure must be relevant to your environment, don’t measure what you don’t need but you must measure everything you do need.  (That sounds like something Confucius could have said …)

From a data governance point of view I like to focus on areas grouped into

  • Data Availability
  • Data Accuracy
  • Data Reliability

Once the dimensions you wish to measure are defined and agreed, you can then ‘KPI’ them, set goals for each measurement, define the thresholds that define success and failure, and then identify ‘owners’ of each KPI.

Defining the owners is one of the important success factors here, ensuring that each KPI is aligned to a relevant owner, someone who is ultimately responsible, and capable of ensuring successful delivery to that KPI is essential.

With clear KPI’s identified and communicated we can start to look at how Continuous Improvement comes in to effect.

Continuous Review

Having structure and defined measurements in place is great, but this is not the end product, the structure needs to define a process and embed a culture of continuous review. Without this continuous review monitoring will eventually cease, causing the improvement process to stall.

I find that this process needs to be “enforced” it is natural for resources to focus on what is new, all the exciting development work. People need to be made responsible for the management of the review process, with set review periods. Resources and their management also need to carry out their own informal reviews on an informal basis. It is a great help to build dashboards and exception reporting to automate as much of the measurement as possible, providing real-time notification when KPI’s are being met or not.

It is continuous review that drives continuous improvement.

Final Word

This is an extensive topic, one that I have only touched on here. I would like to hear your stories on how you are or have implemented a continuous improvement culture in your data management organization. Feel free to let me know.

I will leave you with one of my favorite examples of where continuous improvement was implemented at the front line:

A large global fast food organization had a ‘Data Quality Ladder’ that matched all their stores against each other based on the level of quality data entered through their POS system. These stores were KPI’d against where they came on that ladder. The result was constant competition to try and get to the top of the ladder.

Continuous improvement at its best!