Over the past few months, an increasing portion of my conversations with customers has been about globalization.  The recent financial crisis has caused many companies to focus more on market diversification. Some traditional markets are experiencing slow or negative growth and some emerging markets have recovered faster than others.  In the global search for customers, crossing geographic boundaries has become less of an impediment and more of an opportunity.

Globalization and multi-country operations bring a whole host of complications ranging from tax implications and exchange rate fluctuations to labor regulations and cultural sensitivities.  However, one often overlooked aspect is the impact that globalization has on your enterprise data management strategy – and particularly – your customer data.

One such organization I worked with recently embarked on a global expansion strategy: A new CRM system with Unicode support was implemented and a data migration/consolidation project was underway. Shortly after the new system went live, matching became an urgent priority.

One of their valued customers from Japan was travelling in the UK and called local customer support: unfortunately the customer was not properly identified and the support provided by the UK team did not meet the customer’s SLA. An audit revealed that the CRM system had not one, but two forms of his name – one in the original Japanese multi-byte character set and another transliterated form of his name in Roman/English characters. Neither form matched what the UK support team used as search criteria to locate relevant customer records.

Japanese name data has some unique characteristics: 4 different character sets are used – Kanji, Katakana, Hiragana and Romaji/English. Kanji – which uses the logographic or pictogram characters from Chinese – is frequently entered into computer systems with no spaces between the characters. Transliteration or conversion into Romanized form introduces more challenges – a single Kanji character may have many different phonetic readings.

For example, the Kanji characters below:

may be read as ‘Suzuki’, ‘Shinrai’ or ‘Shinku’.  Similarly, one Kana can have multiple Kanji translations.

is the Hiragana representation for “INARI” but among its various representations in Kanji are:

It all depends on whether you’re talking about a fox [the first in the list] or a person’s name [the rest]

Making all customer data accessible to all regions introduced yet another challenge: Data was being entered by individuals from different cultures, with very different translation and spelling conventions.

For example, the Russian name below is transliterated, spelt and typed differently in the following countries:

Russian: 

English:            KHRUSHCHEV

French:             KHROUCHTCHEV

German:            CHRUSCHTSCHOW

Dutch:               CHROESJTSJOV

Italian:               KRUSCIOV

The IT team at our customer overcame these challenges by implementing Informatica’s Identity Resolution - and delivered a truly global customer search that provided consistent and reliable results regardless of character-set, language or encoding.  The business is now able to reliably identify customers across the globe and deliver the appropriate level of service – thereby increasing renewal rates and upsell opportunities.

Matching technology can help deliver the promise of a global customer view and all the attendant benefits of market and revenue expansion, operational efficiency and customer insight.

However, globalization also increases the importance and complexity of risk management.  In the next part of this series, I will explore how risk management and compliance are affected by customer data that knows no boundaries.