If CRM has been a struggle or a passion for you then Infosys’ CRM blogs is the place to be in. Come join us as we discuss the latest trends, innovations and happenings which will have a bearing on CRM.

« Can Remote Device Monitoring benefit Field Service organizations | Main | Dreamforce 2012 - Ideas Unlimited »

MDM and DW

One of the pertinent questions other than "Why MDM" that comes up quite often in discussions is "how MDM (Master Data Management) is different from DW (Data Warehouse)"?

I think to a certain extent the reason for comparing and thinking MDM in line with DW is in the fact that in consolidation style of MDM, data from different sources is fed into MDM data store and many implementations involve typical ETL for this, which is similar to what happens in DW. Also typically de-duplication, standardization happens (or at-least used to) in DW and the same also happens (and more) in MDM now. In a way you can say MDM is an offshoot of DW and as the famous saying goes "One size doesn't fit all" MDM has evolved as a separate domain on its own. But if we see the typical business use-cases of MDM they are different from DW.  

Business expects complete, consistent and accurate data from an MDM system whereas DW is expected to provide aggregated and historical data for business to take decisions and for reporting purpose. Transactional as well as master data would reside in DW and needless to say only master data would be stored in MDM system.

The Data Warehouse is an OLAP system and MDM is an OLTP system. Well isn't that a sharp contrast that differentiates MDM from DW. To further expand OLAP systems are meant for different things and also are setup for sequential I/O with large memory/buffers for running lengthy queries(involving aggregations) on big data sets and relatively less transactions whereas on the other hand OLTP systems typically involve random i/o and high volume of short on-line transactions (CRUD operations). MDM will be modeled after 3NF (usually) and will have current and up to date data whereas DW will have historical and aggregated data typically stored in multi-dimensional model (star schema).

MDM doesn't store transaction details and hence won't grow as big as DW. Data from MDM can be fed into DW as the data is more trustworthy and reports/BI would get a huge boost from it. Typically data from MDM fits into the dimension tables in a DW. (And transactional details go into fact tables in DW)

DW won't be interfaced with B2C systems whereas MDM systems are increasingly being used in B2C landscape where MDM's Data store would be the system of entry for customer and product information. According to Gartner organizations can reap the best value from MDM when implemented it as an operational system. SOA, BPM are few of the other prominent things that are increasingly becoming part of MDM landscape (either in-built or via integration).

Customer satisfaction, regulatory compliance, cross-sell/up-sell, Mergers & Acquisitions, Campaign effectiveness are some of the top drivers from MDM. Analytics, forecasting, reporting are the business drivers for DW.

To conclude MDM and DW address two different sets of business use cases for an organization and they complement each other.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.

Subscribe to this blog's feed

Follow us on

Blogger Profiles

Infosys on Twitter