For BI to be effective both structured and unstructured data are required. Because so much information resides in emails and other documents, the only way to get a full view of what is happening within a specific business unit or the organization as a whole is by combining financial and other performance related data with documents, geospatial data, and other types of information. Consequently more BI solutions integrate unstructured options into the available types of data capture. This leads to greater diversity in the types of analytics that can be conducted and in turn helps increase the variety of analytics.
Beyond GIS and general mapping, identifying customer sentiments from notes sections are becoming more important as businesses look outside of the organization to identify what is being said about products and services within social networking channels. In addition, the combination of external and internal customer sentiment sources create a more balanced view about what is happening in relation to overall customer experience. Industries requiring heightened sensitivity to potential fraud from both internal and external sources can identify trends in submitted documents, email messages, and general content management systems. No matter what the application, the reality is that as BI use in general matures, so does the way in which businesses look at all areas of analytics and the types of data that can be integrated within the BI infrastructure.
Before organizations expand into unstructured data analysis they need to make sure that their current BI infrastructure is actually working for them. Once businesses decide to take the plunge, many considerations are required. The general considerations to look at when adding unstructured analysis within a structured environment are broad. These include customization, integration, development, etc. In essence these are the same requirements for developing a traditional BI or general analytics solution.