One key element of any successful CRM project is the ability to migrate data and keep data integrated between applications.In order to have a stable and viable CRM environment you need to ensure that all your data integration and migration is done using a supported method. Directly modifying data in the database is not supported for a number of architectural reasons. This means that you should not perform direct inserts, updates or deletes. Instead, you should interface with CRM by leveraging the CRM web services and supported methods. There are different supported ways to read data from CRM: primary one is to use the CRM web service, the other choice is to use SQL statements to query the CRM system.
Since there would be significant development effort to build code from scratch to use the CRM web services, and others have already created integration applications, most choose to use one of the available tools. Some ETL vendors have been adding CRM webservice adapters and other supported tools to their own suites.
From a feature perspective all data integration solutions have some similarities and generally include the following basic features:
- Source-to-target mapping
- Data manipulation and transformation
- Scheduling (for batch-based processing)
- Error handling
From a performance perspective, all data integration solutions are designed to be able to scale by taking advantage of the server computing resources. However, your choice of adapters might perform differently depending on environment.