Blog post

Master Data Management pops up in strange places

By Andrew White | October 07, 2008 | 0 Comments


Two different ideas crossed my mind last week that became connected (with MDM).  These ideas might not seem that related to each other, let along MDM.  One concerns email and contacts, and the other the financial and economic woes we see in the press.  First, a story about email.


Being away in the UK for a few weeks meant that I had to relay on my PDA.  Being so far away from home base meant that I had to trust that the data I had on my PDA, in terms of emails and contacts, was up to date.  It so happened that I received a text from a trusted colleague but that this person’s contact details were not already known to my PDA.  To ensure I could see the name correctly rather than the phone ID I entered the phone ID into the contact data base and keyed in the person’s name.  However, I “fat fingered” the name incorrectly – little did I know at the time.  Next day more emails came in and I noted the incorrectly spelled name.  After a couple days of this I went back into the contact database to correct the name.  I then noted with confidence the new emails came into my PDA with the correctly spelled name.  However, the previous emails, still in my inbox, preserved the incorrectly spelled name.  This is a classic example of MDM failing due to a design principle in my PDA: the link between contact, a master data object, and email, is not preserved: it is only validated at the time the email is received (or sent).  The implications, if I extrapolate to the business level, are that I had created two discrete silos of information that were really for the same relationship.  If this had been more about financial transactions such as invoices, and if I had to pass these emails (invoices) onto other people to execute, I could easily get myself into trouble.


Now for the other story: While I was merrily catching up with email last week I was, probably like you, watching with awe the US and European news channels explaining the woes of the global financial crisis.  One particular story caught my eye, in the legacy communication source known as a newspaper.  I was reading the Financial Times October 5th and there was an article about Glitnir, the (then) third largest bank in Iceland.  Being caught up in the credit crunch, the government of Iceland had decided to nationalize the bank on September 29th.  The reason given was that confidence in the bank had fallen significantly because it was unclear where risk resided with the bank.  The bank had complex and intertwined relationships with numerous other banks and commercial businesses that resulted in it being an unknown risk to many other counterparties.  The government of Iceland had to remove the risk quickly in order to stabilize their own financial system – hence the nationalization of the bank.  The newspaper version of the story in the Financial Times had a wonderful graphic showing the spider-leg like connections between Glitnir, and all manner of European banks and commercial firms, sometimes related to each other in overlapping relationships. 


So why are these two stories related?  Just imagine if the links between Glitnir and its business partners were not preserved accuratelyy?  What if the same design weakness in my PDA had instantiated itself in weak process integrity in the banking system such that firms had no clear definition of who related counterparties were?  What if the names where spelled incorrectly? 

Comments are closed