Information Infrastructure Lifecycle Supporting Business Analytics
by Fern Halper and Marcia Kaufman, Partners
Last month Hurwitz & Associates began a series of articles focused on business analytics technologies and how they are used to support business decision-making. These technologies range from business intelligence and predictive analytics to CRM and SCM solutions. Business analytics are predicated on a solid underlying information infrastructure. One mistake we see companies making is to consider business analytics in isolation from this information infrastructure. The nature of the overall business information lifecycle demands that companies understand the interrelationship between data, data quality, data management, and business analytics.
The Information Infrastructure
What is the end goal for a company’s information infrastructure? Organizations need to be able to dynamically access and leverage clean, reliable information – in order to be more competitive. What is required to do this?
? Predictable Data Quality and Integrity. Even the most sophisticated analysis combined with a sharp, well-designed graphical presentation will provide little benefit to the business decision-maker if the data is of poor quality. However, a large majority of companies do not trust the quality of their data. For example, in a recent Hurwitz & Associates study on SLA implementation, only 27 percent of the respondents felt that their data was highly accurate. This is consistent with other studies that have found that companies, in general, don’t trust the quality of their data. If a company can’t rely on its data, its credibility falls apart. Software solutions for data quality generally profile, match, and standardize data before it is stored. An integral and critical part of having “good” data is the ability to have a correct, clear, and concise view of personal identifiable information. Software that addresses this issue takes data quality to the next step by providing a single view of the customer, partner, or supplier.
? Getting the Data to the Right Place in the Right Way. Once a company is confident in the data they will be feeding to their business analytics software, they must get the data where it needs to go. The next step may be to extract the data from the source, convert it from its previous form to a new form so it can be put into another data store or application, and finally load it into that data store or application.
? Managing Comprehensive, Highly Distributed Data Sources. Typically, data is stored in multiple sources throughout a company, in such places as databases, data warehouses, and even data warehouse appliances. Mapping meta-data, in context, is critical especially in a highly distributed environment. A meta-data management capability insures that the meaning of data across sources is consistent. This meta-data management capability is significant in industries like financial services where rapid, accurate exchanges of meta-data are critical to daily business transactions like financial trading.
? Audit and Verification of Data. Although we put this last, we are not implying that auditing should be done after the fact. Rather, this loops back to the question of data integrity, that is, data that would pass an audit. These issues are particularly important in today’s compliance-driven culture. Software that provides an organization with a view as to who created, changed, deleted, or accessed data can help companies in this regard. Ultimately, a company should be able to monitor its systems in “real time” to look for suspicious activity.
Next month we’ll begin to do a deeper dive into these technologies and spotlight vendors in this space.