Blog post

The Case for Stewardship versus Governance

By Andrew White | June 20, 2017 | 0 Comments

Metadata ManagementInformation StewardshipInformation Governance

This blog is written and focused on “information governance” – and not focused on “IT governance” (governance of technology), or “corporate governance”.  Our most recent effort at reconnecting what we mean by “information governance” and what IT does is captured nicely here in figure 2: Enterprise Digital Governance: Resetting Governance for the Digital Age.  In a nutshell, “what” is to be explicitly governed spans people, process, data and technology.  We no longer hide data inside technology or “IT”.  And by adding ‘data’ as an explicit object, we are expanding the factors of production with a new language: see Information as a Second Language: Enabling Data Literacy for Digital Society.

So for us, “information governance” is a catch-all term that covers the scope of all kinds of data including structured data, unstructured data (as in content), documents, images, analytics, models, rules and so on.   And if you could not tell from our published research, we say “information governance” as opposed to “data governance” since the vast majority of clients we bump into use the term “data” at a more detailed, technical and so IT level.  When we say, ‘information versus data’ to denote a different, more business oriented approach, it starts the conversation off on the right foot.

So that is the basics out of the way.  Now for the point of the blog….

We have said very regularly in the last 8 years or so:

  • The act and work of (information) governance focuses on the setting of (information) policy
  • The act and work of (information) stewardship focuses on the enforcement of (information) policy

Many organizations have set up governance boards and the like; those firms in regulated industries tend to be quite good at this.  But in most cases all firms have struggled with what we might also call the operational end of governance – the enforcement of policy.  I think there are different kinds of enforcement:

  1. The kind IT can do by processing rules and programs to automate how data is cleaned and updated following agreed goals and agreed policies
  2. The kind that business users do when addressing business outcomes, processes or decisions that are held hostage to data

The former work is what IT should be doing all along.  It’s the realm of “data quality tools”.  The latter work is very new, really, and not many firms are good at it.  In fact five or six years ago I would attend our Master Data Management summits and sit and have dinner with clients.  Some of these clients would be large ERP users and they would be struggling with enforcing policy of data used in their business applications.  Fed up waiting for app vendors to bring solutions to market they would build their own.  It is the recognition that this work should be explicitly defined, officially a role (not full time), and supported by new capabilities, that goes a long way to explain the excitement around (information) stewardship.

So right now we are tracking this new segment of information stewardship solutions.  But looking at the bigger picture or broader topic – one of market structure and how it will adapt over time – is quite complex:

  • To what degree can a “metadata management suite” be used to “build” your own information stewardship solution?
  • What is the difference between a tool-set (e.g. metadata management tools) and solutions (e.g. prepackaged with a specific application and role in mind)?
  • Is the market boundary or definition of a tool set the same or different to that of a solution?
  • Will a separate market emerge to support the business work oriented around policy setting and analysis from the metadata management tools market, or will it emerge “on top” of the information stewardship solutions segment?
  • When, and in what form, will data lake, data warehouse, or analytics apps vendors “play” well with others in the area of governing and stewarding data (and analytics)?
  • When does any of these questions require an answer?
  • Are we not paid the “big bucks” to work this stuff out?

Finally, metadata management tools themselves can be used to assemble or build many different and sometimes related applications or solutions ranging from data catalogs, glossaries, dictionaries, business rules libraries, lineage analysis, and many, many more.  Deb Logan and I tackled this a while back – but this note was not very popular with clients: Operationalize Information Governance With Technology Support.

At this point we have our thinking caps on.  To be frank, they are on most of the day.  Whether we come to agreement on the questions, let alone the answers, is another thing.  What do you think?

Comments are closed