Blog post

The Three Rings of Information Governance

By Andrew White | June 11, 2013 | 2 Comments

Master Data ManagementMaster Data LifecycleInformation PolicyInformation ManagementInformation Lifecycle Management (ILM)Information GovernanceInformation ArchitectureInformation Trust

I had a packed call of customer phone calls today.  It was, as usual, exhilarating.  It happens every time – when you cram in a lot of customer interactions into a short time frame, it generates both patterns, and lots of ideas.  It’s what makes being an analyst fun.

The pattern that emerged today I wanted to capture with my catch-phrase, “the three rings of information governance”.  Five times today (out of 7 calls) I asked the customer to draw, in their mind, three concentric circles (see diagram below).  This three ring model is what I used to answer a range of questions – that all ended up at the same point.  How should I think about where and how to govern my enterprise information?

The three rings describe the different degrees to which information is governed in support of improve business outcomes.  In other words:

  • The inner ring represents the master data (fewest number of attributes) that is used most often by the primary/most important business apps/process
  • The outer ring represents that data that is used by the fewest number of applications/processes (i.e. 1) by, ideally, 1 user
  • The middle ring is everything else- and where most of the pain persists – much data that is added to master data, used by a number of business processes/applications, and a number of users

The rings denote the depth and completeness of the information policy decisions and challenges firms face.  I think ALL IG policies can be thought of as different across these three categories of information – including security, quality, trust, storage, etc.

Most organizations get the inner ring – at least they say they tend to obtain broad agreement by most of their business stakeholders and architects.  This is, in reality, where MDM got started.  By managing the common, the core data between the most important processes, information re use can be increased and thus the value of the data increases and investments in programs that use that data are more likely to achieve their business outcomes.

Many organizations also get the outer ring too.  This seems obvious, even if it is hard to implement.  Data in this outer ring was like a quite summers day, floating along on its own merits, not interfering with anyone.  This data does not need to attract the same level of governance that data in the inner ring needs.

The problems are in the middle ring.  This is the realm of all that other data that powers or that uses manage in CRM, e-Commerce, ERP, Supply Chain or any other business application – packaged, developed or otherwise.  This is where the arguments remains today.  We all know the data should be governed – less so than master data and more so than the “edge” data in the outer ring.

We really need all the business applications to play well with each other (fat chance).  There is just so much limited sharing of the base metadata  -that would help with consistency in business rules enforcement, that could sustain a more mature information governance program across the three rings (and all the applications).  Many metadata management vendors just don’t even know how to sell this concept – being happier to sell to IT that needs a data warehouse stood up.

But the point is this: think of these three rings – and think where you would place each attribute in question.  Use it to communicate more clearly what is master data – and what is NOT master data.  Seek to govern that which is most impactful of the business first.  Worry about that other stuff later – much later.

I drew this 3 ring diagram at our recent MDM Summit and my (no retired) friend and colleague Michael Blecher actually used a variant in one of his metadata management notes 2 years ago.  I think it’s a simple model – and I understand it over simplifies the challenge.  But I think we need simple ideas in order to unravel the complexities we face.    It maybe that reality calls for 5 layers – but maybe the 3 layer model conveys the point – and work we have to do.

What do you think?  Does my pattern make sense?  Does the model help?  Would love to hear how others tackle this classification problem.

Comments are closed


  • Angie Tarasoff says:

    Interesting idea – and one that made me wonder about how this integrates with governance of pace-layered applications.

    I’m thinking of taking this idea and marrying it to pace layers, to help my stakeholders leverage prior learning rather than introducing a net new model for thinking about this stuff every time out.

    I need my stakeholders to understand how these pieces fit together into a larger whole. Using a similar frame helps them see the connection between the data, the applications, and the policies – and then figure out how best to allocate decision rights.

    Specialization can work against us – particularly at the strategic level. As I see it, we have to try to simplify or unify models to help leaders at all levels understand how things fit together, rather than looking at each part in isolation of the others.

  • Andrew White says:

    Hi Angie, Thanks for the post. I would agree – it is great to leverage past learning. I don’t actually believe my 3-ring idea is that new (nod to Bill Blitch again) but I do think that refreshing old ideas, and keeping them simple, is helpful.

    I also like your idea of applying the 3-ring idea for information governance to application pace layering. I don’t think each ring applies to the a layer; I think that at each layer, the 3-ring model can be applied and a different set of data would be classified in each of those layers. So you could draw the 3-rings over (or under) the 3 horizontal pace layers.

    I would love to hear if your users like and can use the idea.

    Thanks again.