The role of a data steward

Not so long ago I received an email which, after giving my personal details, allowed me to download a Gartner report. The report talks about the difficulty organisations have understanding and explaining the role of a data steward.

The author starts with mentioning some misconceptions around the role of a data steward, like:

  • It’s a full-time job;
  • It includes data entry;
  • It’s something IT or the shared service/data management office “does”;
  • It’s maintaining and sustaining a master data management (MDM) app or data quality tool.

The report then continues with the fact that not all business users of applications are information stewards. But information stewards are “chief problem solvers” and so are often the few power users of applications. It concludes with, in summary, the role is part-time, tasked to line-function business users or application users (not IT or a shared service or data management office). As such, think of the role of information stewardship as exception-based. Only the really challenging exceptions are routed to the information steward for resolution. Because it is only the most challenging exceptions, it should not take more than a few minutes a week to resolve. If it does, the policies are likely wrong and need to be looked at, or the tools employed by the information steward are not efficient enough.

I have to say that I see this differently.

Data stewards stand with both feet in the business

As an MDM consultant, I have seen quite a few ways that companies implement Data Stewardship. While implementations tend to differ in urgency, scope, beliefs and readiness, to name a few, there are also some resemblances. I’ll summarize some of the resemblances I’ve seen in the implementation of the role of the data steward below.

Data stewards stand with both feet in the business, they’re not in IT, and hardly application power users. They might know a certain part of the application pretty well because they work so much with it. The role of a data steward is to support business requests around master data. IT Support is there to assist with questions around applications.

To a certain extent, it will depend on the size of the organisation and the amount of master data requests whether there is a full-time role for a data steward. As soon as it is possible though, please do make it a full-time position. The more requests that go through the hands of the data steward the better. By seeing more data the data steward becomes more knowledgeable and it will be easier to spot anomalies in future master data requests. For this same reason, funnelling master data request through a shared service centre is a great idea. This will enable data entry to be consistent and of higher quality than when over multiple disparate data stewards or individual users are creating data.

Data stewards are not the ones maintaining either an MDM application nor a data quality application, as that is a typical IT support role. Data quality reports do often go to data stewards. Solving the issues mentioned on data quality reports should never be the role of a data steward. Diminishing the number of data quality issues is to a large extent about training and education. Data quality issues should go back to the data providers to review them and provide correct input. Having data stewards create and maintain master data will automatically lead to less data quality issues. As data stewards are aware of the data quality rules and can review incoming requests.

Remember that MDM is not the end goal, rather a means to an end. The end goal is having correct data, of good quality, that can be trusted, reflects the state of the business entity in real life and is available when needed.

Data stewards are of utmost importance to this process.

They are the ones that make sure the data enters the system correctly and add value to the process. You can start appointing data stewards now if you haven’t done so yet. There is no need to wait till an MDM application implementation.