Within recent conversations with our clients, I have noted a pattern that is concerning to me. Many of these discussions focus on three topics, usually in this order:
Those of you considering implementing an MDM system that makes available integrated master data in a single entity or, more likely, a multiple entity data hub need to give careful consideration to whether that MDM data hub is simply a system of record or whether it is a system of record and also a centralised master data entry system. Those of you considering the latter need to fully understand the impact of this on your enterprise so that you can budget and plan well without surprises. - Read the rest of this entry »
Following on from my last blog, once you have common definitions for master data, the disparate data sources and the mappings you can then profile the data in those disparate sources to understand the quality of the master data subsets there. - Read the rest of this entry »
I was reviewing some material I had written about 1 1/2 years ago on MDM and it struck me that I have gone through some revisions on my opinions regarding the definition of master data. At first my perception was that master data, by definition, should be extracted from its numerous source systems and consolidated into its own data repository, to be accessed by downstream applications. - Read the rest of this entry »
There seems to be a lot of debate around about data governance and MDM, return on investment on MDM etc. going on at present. However the discussion is still a data oriented and IT oriented one with several of my clients still asking for help on MDM business cases at present. So I thought I would look at what MDM could potentially do for an enterprise and see if it implementing MDM might have wider implications. I was amazed at how long my list of what MDM could potentially integrate with started to get. - Read the rest of this entry »
I just reviewed a handful of case studies on master data management, and I had the distinct feeling of deja vu. Many of the MDM programs in the case studies centered (of course) on customer data, and even more pointedly, on the matching and linkage aspects of customer data integration. Considering that five years ago (prior to the creation of the term “master data management”), these case studies would have been touted as best practices in data quality. - Read the rest of this entry »
In one recent post I noodled on the concept of master data synchrony, and promised to follow up on some dimensions of synchronization that may drive application requirements that feed the decision regarding the underlying MDM architecture. The proposed dimensions are: - Read the rest of this entry »
It’s that time of year again when predictions are made about the coming year. So I thought I would add my two cents worth before getting back into more focussed blogging this year. So here goes: - Read the rest of this entry »
What is the qualitative difference between the traditional manner that data quality tools are used for data cleansing and master data management? The answer is related to the way that business application owners have specified their needs for consistency and synchronization of the critical data assets for both consolidation and use. - Read the rest of this entry »
When I look at the almost every issue associated with data I almost always end up at the same starting point. Metadata. Don’t you hate that word? When I first heard it all I could think about was 60 shades of grey. I got the usual definition - you know the “data about data” one. I struggled with that so I started to break it down into smaller pieces. In the end I came to one fundamental question. “When it comes to data, does the user know what it means?”. If I fail that test then surely the rest may as well not matter. It’s like languages. The people of the world speak many languages all of which are formally defined. Having formal definition makes it understandable. - Read the rest of this entry »
The blog content appearing on this site does not necessarily represent the opinions of DataFlux
