Data Matching Interfaces


	Data Matching Interfaces Many companies find that they have invested heavily in Customer Relationship Management or Marketing (CRM) and have found either there was considerably more cost in the loading of their data from existing systems or bringing in new data problematic. Mainly this is due to the problem of matching data together, and because data is stored in different formats, it need some sophistication in the data handling and matching techniques to get the required consistency in the data, which the CRM system does not have. This is an issue many CRM systems fail to address.


	Diverse Data When bringing diverse data sources together, or even common data sources, it is not easy unless you have unique references (reference numbers, account numbers or policy keys) linking the data together. If there are no unique numbers then this normally leaves name and address or other unique data fields which identify individuals that can be matched together.


	Process To start the matching data without reference keys, first the data needs to be disseminated into constituent parts, so that common parts can be identified and readied for matching. The matching can vary for the different fields and the methods are exact whole field matching, part field matching, Fuzzy matching, PWM (percentage word match), Soundex and Metaphone matching. These routines can be built into single programs, can be part routines of other programs, can be stand-alone with GUI front ends or can be executed on the fly within online systems.