Enterprise data integration can be performed in many ways. One common method is to build a central corporate data warehouse and then funnel in information from various points within the network or from other business operations. For instance, data could enter the data warehouse through a set of middleware applications or through human data entry. Staffers might take data from mailing lists or written customer surveys and integrate them into the data warehouse. Moreover, data may come in automatically through software setups, for instance, from a customer's mobile phone or from a call center environment.
One of the issues with enterprise data integration is that companies tend to have different kinds of data in radically different formats. Part of enterprise data integration and enterprise data management is handling the sets of unequal data. Different kinds of data include structured data, formatted in easy database table designs, and unstructured data, which is often composed of raw text or raw data sets that are not formatted for database use. An enterprise data integration plan often provides solutions on not only how to integrate all of this data, but also how to make it readable and useful for the business. New technologies accommodate more handling and manipulation of relatively unstructured data sets.