Gartner Blog Network

The Tangled Web Systems Integrators Weave

by Andrew White  |  August 26, 2014  |  8 Comments

I was on an inquiry today with an end-user.  It went like this:

  • “We have an ERP system –its pretty much stable and OK – but the data within it is not really at the level of quality and consistency we need.  We thought we needed to adopt MDM.
  • “We have new and emerging requests to clean and govern that data, or some of it – along with other data – used and likely to be used by other applications and even our customers
  • “We have a Systems integrator advising is “not to worry about MDM” per se, and instead to “do a complete new ERP replacement since newer ERP systems are better at doing MDM”

Really?  Um no, that’s not true.  That is mixing metaphors.  Why use a hammer on a screw?

ERP systems were NEVER, and are NOT NOW, designed to “do MDM”.  They might today do a better job of managing their own (application specific) data – but that is NOT the same as MDM.  MDM came about due to the need to govern information (for re-use) across multiple applications or processes, whatever they are called.  There is a certain overlap or similarity, but the two efforts are quite different.  The ERP focus is ERP – so managing data in the ERP system.  And MDM focus is not ERP focused – it spans ERP and any other application that published to or consumes from, ERP and any other application.  Note that MDM is not “wall to wall” either – whereas ERP was once part of that sales pitch!

I get the feeling the System Integrator was selling this end-user “the old ERP road map of the 1980’s”.  This old model would go like this:

  •  Sells ERP for a large amount of money
  • Get to “user acceptance” as fast as possible *
  • Get payment and get out of there ASAP
  • Go past “go” and get paid again: Get called back, typically 9 months later, for another gig to “help clean up the data and ‘do’ that information governance thing”.

* Software and consulting vendors have little interest in selling you the ideal, up to date, road map for ERP that INCLUDES the needed information governance to include “before ERP, across ERP, and after ERP” since they are trying to sell THAT at a higher price than the original ERP footprint (w/o information governance).  They would much rather you come back for more – and keep feeding their bank account.  In a fair world, ERP would be CHEAPER, and part of the difference in cost would cover the MDM and information governance part.  But that, my friends, is all in the art of the negotiation.

Here’s the kicker: though I refer to ERP, any 3 letter-named application suite counts just as well here – CRM, SRM/Procurement, SCM and so on.  Even “application developed” and industry-specific application often works as well.

Category: application-development  business-applications  crm  customer-data-integration-cdi  erp  scm  srmprocurement  

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

Thoughts on The Tangled Web Systems Integrators Weave

  1. Katherine L says:

    Great blog, Andrew. When you talk about master data being used across multiple systems, would you consider (for example), header data in SAP that is used across mutiple modules to be master data or application data? (e.g., company codes, functional areas, cost objects)

    While SAP itself isn’t designed to “do” MDM, it does have a tool that can help with MDM (Information Steward). What are your thoughts on ERP-offered MDM tools that can bring in data from many applications?

  2. Andrew White says:

    Hi Katherine,

    Thanks for stopping by and leaving a comment, and question too.

    We would actually classify company codes, cost codes, and other “look up” tables as mostly reference data. That is data that looks, smells and behaves like master data but tends not to ‘describe what a company does’ (that things like “customer” or “product” do). Most MDM programs include an amount of reference data – you can’t do MDM of product data without a list of permissible units of measure. And some organizations actually start their efforts with a reference data management program before tackling the ‘larger’, more politically charged data like master data.

    On your second point, yes ERP vendors or more generally application vendors can help in an MDM infrastructure. SAP Information Steward, originally designed for use with their own data warehouse, is a great example. It has probably more relevancy, in my mind, in the operational/business apps world than in a downstream data warehouse (even though it can still help there).

    Thanks again and I hope the response makes sense.

  3. Katherine L says:

    Thanks! That does help. Sounds like you could categorize the reference data as “company metadata” (data that helps categorize intra-company information), whereas true metadata would, as you say, “describe what the company does.”

    So, putting together a list that summarizes the categories of data companies see, it would be something like:
    – Master Data (crosses applications, describes what the company does)
    – Reference Data (crosses applications, desribes the company)
    – Application Data (within an application)
    – Metadata (application specific and application agnostic, describes the data itself)

    Do you have any recommended reading re: master data topics? (Checked out the Magic Quandrant for Data Quality already)

  4. Andrew White says:

    Hi Katherine,

    The way I would list the categories is as follows:

    – Master Data (crosses applications, describes what the company does)
    – Reference Data (crosses applications, descries commonly referenced and resources codes and information)
    – Application Data (within an application, not reused outside it, and linked to master data and reference data used in the application)

    And then spanning all three:
    – Metadata (describes the data itself, rules, and polices etc., useful “in application” as well as “between applications” for broader use)

    What do you think?

  5. Katherine L says:

    Makes perfect sense. Now how to work this into the common lexicon?

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.