Gartner Blog Network

SAS’s Definition of Data Steward – Does it Help or Hinder? I think it is Confusing…

by Andrew White  |  September 1, 2016  |  Comments Off on SAS’s Definition of Data Steward – Does it Help or Hinder? I think it is Confusing…

Never to be known to shy away from reviewing a vendor marketing piece without being asked (I used to work for a vendor, too!) I have to comment on what at first blush looks like a “must-read”.  I saw this headline, “5 Models of Data Stewardship” only to be sucked into a vendor marketing paper.  Well, I read it, since “stewardship” is a hot word right now.  I have to say I am glad I read the note.  I disagree with the way in which the term stewardship is used and unfortunately the note explains a lot about why clients are so confused, and as such, why so much information governance goes wanting or failing.  First some issues and then some praise, and I end on an answered question posted by SAS’s contribution to the topic.

Try these quotes for size:

  • “Data stewardship assigns decision rights around data – enforcing accountability”.  No, sorry, information governance assigns decision rights and sets policy.  Information stewardship only enforces policy and assigns no rights at all.
  • “Whether they exist in the business or in IT,” Huh?  Information stewards only reside in the business – it is a business role, not an IT role.  Another classic error.  This makes me think that the vendor is marketing “steward” but they may mean another role
  • Apparently there is such a thing as a “data stewardship program”.  I don’t think so.  It’s a role that is a critical part for operationalizing the work of any information governance program, including but not limited to, MDM

Here’s an example of clever vendor speak (I know, I used to use the same when I was a vendor):

“Indeed, some of the most effective data stewardship programs we’ve seen have bucked popular dogma and started in IT.”

I am sure the vendor has seen some success often led out of IT, but it was not with the role of steward; it was likely some other role doing something with data quality that led to some success.  But to claim it was some kind of stewardship is to confuse everyone.  It’s a bit like looking at one animal, say a lion, and suggesting that in reality it’s a goose.

The paper does however do a great job in helping you determine if you are ready for an information governance program:

  • What are your data skills?
  • Culture challenges?
  • Reputation of data
  • Views of data ownership and more

But just to confirm how we can all bamboozle each of us with fancy phrases, try this one for size:

“As we’ve worked with clients to formalize data governance programs and to institute data management best practices”.

Now I am not sure what it is we are talking about.  First stewardship sounded like governance (assignment of rights), now the paper refers to information governance, and then also data management.  Maybe we all have the same idea of what each of these things are – who knows.  But let’s leave that for now.

So let’s explore the 5 models the paper calls out:

  • Data Stewards by subject area (such as customer, or party, or product or thing).

Yes I can agree with this – we have seen many of these.  This is practical.

The one thing I didn’t like is the graphic – there was “IT governance and corporate governance driving data governance” and no representation of “business” or operations.  Corporate governance is too high level, and IT governance has little to “drive” information governance.  At least that has been an ongoing challenge for years.  It is only recently that we have started to consider “IT governance” together with business governance – but we would not call that IT governance – we now call this enterprise digital governance.  But that does not exist out there…yet.  To be fair the wording that goes along with the graphic makes it clear that corporate governance really means business and operations – but for many firms, these operate at very different levels.

Finally something I can agree with:

“The data governance process will involve data stewards (who are participants), but ulti­mately data stewards will be directly accountable for the success of the management of their data domains.”

Yes, but this was not mentioned in the introduction and neither did the vendor explain the relationship between governance and stewardship.  Also, if stewardship is part of governance, why can there be a “stewardship program” as noted above?  Can you have stewardship without governance?  No.

  • Data Stewards by function (such as customer service, supply chain).

Yes, again I can agree – we see lots of this.  Nothing else to add.

  • Data Stewards by Business Process.

This we have seen but most often it’s a process owner that takes on the responsibility for data.  However this has not been a model that seems to last long.  The main reason is that the most valuable data is most often shared or re-used, it does not rarely remain solely within a process.

  • Data Stewards by System.

OK a logical model – not one I personally have seen and nor would I expect to.  Stewardship is a business role and I don’t know many business people that could logically enforce policy at a system level.  I guess it’s possible but I have not seen it.  Few business people operate in this format.

It is also worth noting the following from our research: models 3 (business process) and 4 (system) are often temporary (but as long as 1-2 years in some cases), as there is frequently an organic (or planned) knowledge transfer to the stewards in model 1 (domain), and after that the stewards in models 3 and 4 are “released.”

  • Data Steward by Project.

This seems to be either a) a data quality project that needs business input, so why not ask a steward? Or b) a one of project that needs to feed in some business acumen so why not ask the steward.

The last part of the paper is, I have to say, very good indeed.  The authors focus on the role as business role (we agree) and they focus on outcomes that drive the work (we agree).  They talk of a formal role with job description (we agree).  They even talk of a day in the life of a steward (we agree on that, since we published that one a few years ago now).   The problem pops up in the scenario of this “day in the life”, and this not a problem for this vendor but for all of us, and the most useful part of the paper.  The steward is none other than a data analyst in the data operations team.  So here is the battle being fought over today in the marketplace:

  • The most successful information stewards have been, to date, those that are business roles as in line-function, working in the context of an information governance program, even MDM.
  • The entire world of business intelligence and analytics, built on data warehousing and now data lakes, have struggled terribly to establish “stewardship” for one simple reason: business people have no interest or incentive, for the most part, to help IT clean up data in “their” data warehouse (let alone the new data lake).  The only time this worked (if work is the right word) is in regulated industry where you had to do this in order to comply.
  • The role of data analysts, business analyst, and now data science, has yet to be reconciled to the role of steward.  One (the latter) is a line-function role, the others are support roles.

There are no answers to this question of how to reconcile the two, whatever the vendor(s) claim.  The answer won’t be found in technology or discovery or data perpetration or MDM.  It is a new battle-front that is only now being recognized.  My blog here may seem to be critical of the vendor paper – and maybe I am being over critical.  I don’t mean to be and I respect SAS and their efforts broadly.  But in the heat of battle or in peace negotiations we need a common language, a common understanding of the core weapons and wishes at our disposal.  If we don’t, we won’t be able to build an effective peace.  The paper is a well intention piece, I am sure, but it does a good job of confusing the market as much as it does give good advice.

PS one reason we don’t say “data steward” and we do say “information steward” is that ‘data’ tends to be the prevue of IT, whereas information (something more meaningful than data) has been known to be of interest to business people 🙂 but this is as much a marketing exercise as it is an education.


Additional Resources

View Free, Relevant Gartner Research

Gartner's research helps you cut through the complexity and deliver the knowledge you need to make the right decisions quickly, and with confidence.

Read Free Gartner Research

Category: data-stewardship  information-governance  information-steward  information-stewardship  sas-dataflux  

Andrew White
Research VP
8 years at Gartner
22 years IT industry

Andrew White is a Distinguished Analyst and VP. His roles include Chief of Research and Content Lead for Data and Analytics. His main research focus is data and analytics strategy, platforms, and governance. Read Full Bio

Comments are closed

Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.