Organizations that focus on individual consumers often to struggle to identify their customers at the most basic level – their name. There are many reasons for this:

• Capture-dependent: spelling mistakes
• Customer-dependent: name changes
• Application-dependent: packing multiple fields into a single field
• Architecture-dependent: conflicting names for the same person across systems

These different types of issues then become increasingly difficult to address in a complex organization such as a retail bank or telco where dozens to hundreds of systems may hold customer records.

Collectively, Customer Data Integration (CDI) means doing all these things well and helps address what was a cause of failure on many Customer Relationship Management (CRM) implementations. Vendors such as IBM, Syperion, Initiate and Oracle offer CDI-specific Solutions and the market is undergoing rapid growth.

Over the last few years there have been significant benefits to addressing these issues through better governance, data quality improvement programmes and upgrades to new applications that were more sophisticated in their capability to store customer data.

This involves fixing historical issues and minimizing the chance of errors occurring in the future.

One of the challenges that globalization brings is around name recognition. Techniques that have been applied over the past few years simply do not work as well with many Eastern European, North African, Middle Eastern and Asian names. The phonetic translations that convert Arabic names into a Western form are typically inconsistent.

Living in London, I see the Retail Banking sector facing perhaps the greatest complexity worldwide. Rapidly changing demographics require new techniques and technologies to solve this name recognition issue. Once again, big vendors are moving into this space through acquisition – with IBM offering a specific product – GNR – to meet the globalization name challenge.

One of the key questions is who should sponsor Information Managements.  The governance sections of MIKE2.0 describes operational organizations and how to get there, but makes the assumption that the CEO, CIO, CFO etc. are supportive of the initiative and will act as sponsors.  What happens when they’re not?

Actually, it seems that this is the case more often than you would wish with many senior executives unwilling to commit to the proper management of information.  It’s not hard to work out the reason why, in most companies (and increasingly in many government organizations) the CEO is only appointed for a short contract with rapid rotation of new talent into the role.  No wonder the CEO acts like a politician looking for the “quick fix” common sense answer that they can put in place within their term and position themselves to be extended (analogous to a politician seeking re-election),

There is hope, however, by looking at the board.  In most companies, board members have a longer tenure than CEOs and also feel more exposed to legal issues.  A quick conversation about the issues of ledger versus non-ledger data (discussed before in this blog) highlights to board members how great their exposure is if they don’t mandate better governance.  Judicious use of passionate middle managers can complete the pincer movement and before you know it the CEO sees Information Management as a mandatory activity and a quick win.

MIKE2.0 uses “small world” measures to assess data models (see Small Worlds Data Transformation Measure). The challenge many users have found is that for large models, it is very difficult to calculate these metrics. Thomas Isaksson has created an open source tool and provided the code on SourceForge which automates the process , supporting any RDBMS for which you have a JDBC driver or ERWin (via CSV output).

We hope that this initiative will further the encourage the adoption of these data model metrics and help demystify the traditional modelling process. Please take the time to use, test and extend this beta code.

Michael zur Muehlen from the Stevens Institute of Technology was recently interviewed as part of the AT&T Techchannel roundtable on Web 2.0. During the interview he used BearingPoint’s Information Management initiative as an example of Web 2.0 in action and directed viewers to our open source site. You can view the program on YouTube (part 1: http://www.youtube.com/watch?v=WIm5txBm1YA, part 2: http://www.youtube.com/watch?v=5it8oBMbPrg).

Even if you have already used MIKE2.0 resources, it is always worth re-visiting as new materials are being constantly loaded. Some examples of new additions to MIKE2.0 include material supporting data mart consolidation, role and skill mapping, eDiscovery, understanding the information implications of hosted CRM, protecting your enterprise content and XBRL.