The analysis of unstructured data is becoming more prevalent as organizations move beyond traditional forms of analyses to gain competitive advantage within the marketplace. Capturing transactional data and sales figures is no longer enough to keep organizations competitive or to maintain customer loyalty and satisfaction. This is especially true as unstructured data represents over half of an organization’s information sources. To stay competitive, organizations are shifting towards the analysis of unstructured data to gain additional insight into their customers, suppliers and competitors. The question organizations should ask when considering the analysis of unstructured data is how does leveraging unstructured data translate into better competitive advantage, increased customer loyalty and improved performance? The answer to this question lies in the identification of how current uses of unstructured data are solving business issues.
This article presents some practical applications of unstructured data analysis that move beyond enterprise search. Organizations can use the examples provided as a springboard to identify how unstructured data can be leveraged within their organization to help address business issues.
The adoption of unstructured data use within BI and BPM is in its infancy. With the introduction of search tools to enable easy access to developed reports and BI-based information, the perceived value of unstructured data is increasing. Consequently, an increase in understanding how unstructured data use benefits organizations exists. The applications of unstructured data analysis below identify how it can be used to identify trends and to increase performance within the organization.
Customer experience management
Call centers collect a wealth of customer information that extends beyond customer-related information and buying patterns. When customers call to complain about bad service, defective products, or other issues, this textual information is collected in the CRM system. Also, content collected from emails and customer satisfaction surveys identify how customers view the organization’s products or services, whether they recommend these services to others, and what can be done to improve customer perceptions. The text based data collected from these sources help organizations understand their customers fully to help create a complete picture of what customers want. This translates into increased sales and customer loyalty affecting the overall bottom line.
Within marketing, there are several applications of text analytics. Organizations can search public information such as press releases and websites to gain an understanding of the overall market and general market trends to position their products and services better. In addition, the information collected to enhance the customer experience can be used to develop marketing campaigns based on customer segmentation. The more information marketing collects, the more likely their initiatives will match the products customers are willing to buy. Examples include retailers’ ability to develop campaigns based on in-store consumer buying behavior, the comparison of customer feedback before and after product launches, and targeted campaigns based on valued customer programs.
In general, there is an overlap between customer-based and marketing-based unstructured data usage. The information collected for each business unit can be utilized for analysis in both departments as customer focus provides the key business drivers within each. Organizations can consolidate their initiatives to gain benefit from the variety of information gathered and used to enhance the overall view of the customer and to target their interactions based on buying habits, lifetime value, and demographics.
Healthcare and patient care
The healthcare industry and management of patient records represents an important way in which unstructured data use helps healthcare professionals and patients. Notes within patient records help doctors and other healthcare professionals identify a patient’s medical history to assist with proper diagnoses of health ailments, prescriptions, and potential side effects based on historical patterns. The potential to diagnose patients incorrectly is lessened due to the ability to create an overall view of the patient as opposed to a diagnosis based on a single visit.
Aside from patient records, digital images such as X-rays and CAT scans can be emailed and analyzed for diagnosis to shorten the wait time to quicken the treatment of potentially life threatening ailments. Additionally, the information collected along with the digital images can be used to identify commonalities in medical conditions and successful treatments.
Unfortunately, fraud normally occurs from within the organization. Insurance claims, mortgages and other financial documentation can be recreated using accepted claims forms or accepted mortgage documents to create fictitious documentation. Both structured data and unstructured data can be collected and analyzed to identify similar patterns within submitted claims. This will help to help prevent multiple claims with a certain percentage of overlap from being submitted and approved without further inspection. Also, because organizations do not always know what information will prove important, the identification of patterns within unstructured sources can help clarify issues that arise within structured data pattern recognition. This form of pattern recognition extends beyond the financial services arena to include government, retail, healthcare, etc.