Thursday, September 22, 2005

Mastering data

At the 2005 Kalido User Group this week in London a survey was carried out of the attendees regarding the attendees persepctives on master data management. The striking result in the survey was that, although around two-thirds of the respondents (and these are serious companies, like BP, Unilever, Philips etc) felt that dealing with their companies master data was a "top three"priority issue for them, no less than 90% felt that the industry had failed to address it properly. While there are a few software products out there to help tackle customer data integration and product information management, very few address general issue of managing master data across a global corporation.

Large corporations are need to manage not just customers and products, but also other data such as brand, organization, people, price etc, which are scattered throughout a wide range of corporate systems, including multiple instances of ERP systems from the same vendor. The application consolidation that has been occurring in recent years has clearly failed to make inroads into this issue in the eyes of the people that matter: the customers.


Blogger Rajeev Kumar said...

Andy, It is refreshing to see a point of view which reflects the reality more than the marketing hype about master data. A brand manager asked me to explain what is "Master Data Management" and
Is Master Data Management something new?

Talking with industry folks, you will realize very quickly that its old wine in new bottle. Companies have long struggled with this issue and have attempted to consolidate systems (remember post mainframe era). So why the buzz?

The answer is 2 folds

1. There is a new wave of consolidation. A large number of siloed data marts have outlived their age and companies are feeling the pain in maintaining them. Hence a new wave for consolidating them to an Enterprise Wide Data Warehosue (EDW). I hate to say this, but those are mostly Kimbal approach implementations done in a half baked manner. After all isnt "Conformed Dimensions" concept same as "Master Data".

2. Mergers and acquisitions of global companies requiring consolidation of multiple ERP/CRM/SCM implementation.

Both of these present a perfect opportunity to tackle the issue of master data management. The question several people ask is "Why would I go for a myopic approach of MDM for just customer or product" as opposed to a general master data management strategy.

The series of product acquisitions done by IBM and Oracle to get in the Master Data Management space is in my opinion a reactive response as opposed to a visionary step.

I firmly believe that it makes a lot of sense to establish a master data management "infrastructure", "organization setup" and the "discipline". The master versions of data elements (customer, product, people or chart of accounts) will naturally follow.

Having implemented Kalido solution at PPC,I would congratulate you for offering a true solution to the industry needs.

Rajeev Kumar

7:34 PM  
Anonymous Anonymous said...

I am not familiar with Kalido Solutions but came across this site in a google search on "data governance". I have been working with companies for several years on master data management "infrastructure", "organization setup" and the "discipline" only we call it data governance and stewardship. I am interested though in your experiences with regard of how to get executives to understand the "big picture" in correlating the "infrastructure", "organization setup" and "discipline" with resolving their information needs.

2:00 PM  
Blogger Andy Hayler said...

I think the question of how to get business execs to understamd the "big picture" is an interesting one. Rather like data quality, MDM is a business issue, not an IT issue. Inconsistent master data can have a lot of real-world impact e.g. deliveries going astray, payments being made incorrectly. Broadly, companies with master data problems will spend a huge amount of energy arguing internally about "whose numbers are right" and may indeed be taking major decisions based on little, or misleading, data. I suggest that an exercise where you try to quantify the impact of poor master on your particular company would help make it more real for management.

8:04 AM  
Blogger Andy Hayler said...

Rajeev's comment about the need for general purpose master data management is a profound one. It is bewildering to me as to why the problem is being tackled piecemeal, through products addressing "customer" and "product". One of our customers has 350 different master data types to manage. Are we really to expect 350 different approaches and solutions?

8:06 AM  

Post a Comment

<< Home