When I think back to the many data defects I’ve witnessed over the years, one of the biggest causes time and time again is conflicting rewards. All the technology, workshops and cultural change in the world cannot hope to turn around your data quality fortunes if you’re not aligning rewards.

In this post I want to give you a simple resource to help you get started with identifying rewards that conflict with your data quality goals.

As a data quality leader or evangelist, it can be incredibly frustrating seeing the same data quality issues created repeatedly. Even with preventative measures and controls, users will still need to enter information manually at certain points in the information life cycle -- and it’s at this stage that the rewards problem creeps in.

Business users are rarely incentivised for the quality of data they create. Instead, their performance is judged by the number of customer queries they complete in a day, the speed of which they complete a customer site visit or other operational targets.

In short, when it comes to rewards and measures, the business generally wins hands-down. So to help you resolve this issue, I’ve identified a few steps that should point you in the right direction:

• Step 1: Identify the information chains and data sources you wish to improve.

• Step 2: Perform a data quality assessment of the information sources and identify any areas of concern.

• Step 3: Locate the business users who are responsible for creating or updating the information sources that were found to contain issues.

• Step 4: Meet with the team leaders of the workers and identify what their performance measures and goal expectations are, catalogue each measure against the user and map each user against the observed defects.

This series of actions will help you create a matrix of reward conflicts against data defects. For example, in a call centre we observed that a lot of records contained incomplete data. In a previous assessment, it was found that call centre workers were incentivised for the number of calls they could complete and the duration of each call. This created conflicts against the goal of improved data quality.

I’ve witnessed field engineers ignoring the collection of vital pieces of data because they were motivated to complete set targets for completion rates. As a result, the quality of information gathered often suffered.

Once you can visualise the impact of these conflicting rewards systems, it becomes easier working with stakeholders to develop ways to target problematic data sources and make life easier for knowledge workers and data quality leaders alike.