Know Your Customer
Wherever She Shops.

How Big Is My Customer Data Problem?

Customer data is the fuel that drives the modern revenue growth engine and the key to building stronger customer relationships – except when the data is wrong. Around 6% of annual revenue is lost through poor quality data, according to research by Royal Mail. Using a data quality tool is “the easiest 5% ROI lift you might ever achieve,” according to the Canadian Marketing Association.

However, Global-Z wanted to determine if it is possible to achieve even better results than the 5%-6% lift that an automated data quality tool alone may provide.

We learned the answer when a large multi-national corporation came to Global-Z to examine its customer list of more than 10-million records. The first data quality audit used sophisticated software explicitly designed to handle “dirty data.” Then, a second audit used Global-Z’s Master Data Quality (MDQ) process, which adds the knowledge of human subject matter experts to improve results.

Phase 1: Initial Data Quality Audit

During the initial data quality audit, Global-Z found significant data issues in the following areas:

Erroneous data in the city, state, and/or postal code fields
Erroneous data in the address line field
Erroneous data in the name fields
Extraneous data in many fields

Automated cleansing, matching, and deduplication yielded excellent results. The chart below shows what the system considers to be usable data, where “usable” data is defined as records either verified as valid or are likely to be valid, but could not be verified.

Related Topics:

Address Cleansing

Personal Names Around the World

Original Records	11,500,000
Identified as Usable	6,235,000
Percentage Usable	55%

The Global-Z Master Data Quality audit combines the automated software tuned to the specific client situation with analysis by subject matter experts with decades of experience in analyzing international contact attributes.

The experts are assisted by statistical processes that highlight potential issues, such as terms in any language that appeared with unexpected frequency. Among the issues identified by the MDQ process in addition to what was found by software alone are the following:

43%did not have data in the country field, which would have made them invalid. Fortunately, many countries could be derived from validated addresses and phone numbers by the subject matter experts familiar with the local markets.
38%had unusable terms or phrases, including terms such as “Needs New Address.” Many terms were not in English but could be identified as unusable using a statistical analysis that applies in any language.

12%had duplicate information in different fields. For example, the same entry appeared in the ADDRESS 2 and CITY fields.
5.5%had identical values in the given name field and the surname field. In this case, these fields should not be used in the matching process as there is likely an error in one or the other field.

The result of the MDQ audit determined the following:

FALSE POSITIVES: Invalid records that Phase 1 identified as usable	460,000	4.0%
FALSE NEGATIVES: Usable records that Phase 1 identified as invalid	149,500	1.3%
Records Changed	609,500	5.3%
New Usable	6,014,500

5.3% of the records were repaired by Global-Z after the automated cleansing process.

For this particular customer, the critical issues were to better model their customer needs and to improve their quality brand image. Without Global-Z Master Data Quality, 609,000 of their 11.5-million customers would not have received the quality interaction this company demands or they would have had no interaction at all.

For other companies, the quality improvement over the automated process can be broken down into both cost savings (from data not used because it was found to be invalid) and additional revenue from found customers (invalid customer data that became usable).

The larger impact is lifetime revenue value from the 149,500 recovered customers that the software-only solution marked as invalid. If just 20% of these customers spend an average of $200 per year, the lost revenue due to these false negatives would be nearly $6-million in the first year with additional revenue in future years.

To learn more about this analysis, the Global-Z Master Data Quality process, or how it could be applied in your situation, please click on the Contact Global-Z button below.