When proposing a data governance programme, a frequent response from people is “but we already have data governance”. Well, in theory, that may be the case, but the reality is sometimes a little different. A business may have a data governance organisation in place, but sometimes it doesn’t really work (either fully or partially). A kind of fake governance. Like one of those old wild west movie sets. Sure, it looks like it’s a real town, but step inside any door and you’ll discover its all just a façade. In those circumstances, what can happen is that people outside the governance organisation begin to unofficially manage things or worse, people just give up.
Gerard Bartley, Independent Consultant
Gerard will be chairing and speaking at the Data Governance Conference & Master Data Management Summit 15-17 November 2021 on the subject, ‘How Can Tools Help Me Manage My Master Data‘ and ‘How to Build, Manage & Enhance a Data Team‘ alongside Ieva Langenfelde, Data Governance Manager, of Heineken.
The conference is co-located with the Virtual Enterprise Data & Business Intelligence and Analytics Conference Europe. This article was previously published here.
So how do you spot if your data governance might not be real? There’s no sure-fire way but here are some tell-tale signs. However, a word of caution… Sometimes the situations mentioned below work just fine. If your all data issues are solved promptly, your data set is well documented, it is all understood, it is clean and is used fully in your business, then great. Move on. But if not, if stuff takes forever to get fixed (or never gets fixed) then the below could be problems. Do you recognise any?
Governance by committee
Is your data owned by large groups of people or by departments? For instance, is your data owned by BPO organisations or PTO organisations or by large operational functions like Supply Chain. There’s nothing wrong with that in principle, but unless you have efficient triage processes for collecting and dealing with issues arising, there’s a danger that stuff just gets lost.
Data is categorised in very large pools with little granularity
Are full data domains or very large objects owned as single uber-items, with no ownership for sub-components? For example, is all of “materials” being owned by, say, R&D with no differentiation between fields inside it. They just own the lot? Again, this can work, however as R&D could then be responsible for fields that are not in their direct remit (such as costing) there could be a tendency to just deal with the stuff they know about first and the rest gets a lower priority (or even gets passed off somewhere else, without proper control).
Lack of clarity over who owns what
There is no clear governance structure available about who owns what, so no-one knows who to go to when something needs deciding.
No documentation/no knowledge
There is little documentation about how the data works or there are holes where the explanation for the tricky bits should be. Governance is all about making decisions but you need to be informed to make decisions. (How do I find the right data owner?) Hence, no-one can govern if they don’t understand what they are governing.
No overall data owner
There is no overall data owner in the business. No single person to drive data awareness and ownership. Many businesses run very well without such a Data Tsar, but the lack of one may indicate a low data maturity.
So if data issues don’t seem to be being solved and you have one or more of the above present, you may have a problem with your governance… But what can you do if you think you have a problem? Well a starting point, as always, is transparency about the situation. Just get the message out there (politely, with no finger pointing). I’ve found it a powerful tool in the past to flag issues and follow that with the suggestion that the fields concerned are effectively “unowned”. Once the issues is recognised, you stand a much better chance of implementing real governance.
Gerard is a leading speaker on the subject of data management and governance, having spoken at and chaired numerous conferences over the last 10 years. A firm believer in the principle that good data management is all about people as well as IT systems, Gerard’s proven approach focuses on delivering results through stakeholder buy-in, organisational change, process improvement and people development. He has delivered large-scale data governance organisations, built data management teams, implemented workflow systems, delivered effective data management processes and built data quality systems for multiple blue chip organisations. He is currently supporting the BMI group on their data management and governance journey.
Copyright Gerard Bartley, Independent Consultant