With the right steps, managers can reap new benefits with forward-looking metrics.
by Suneth Jayawardhane, May 2013
Analysis of historical data has been the staple of traditional business analytics, in which reporting solutions are used to generate dashboards and standard and ad hoc reports to gain visibility into past and present business performance. Despite the success of these solutions, executives are always looking to gain better insight into the future.
Recent growth in available data and statistical analysis tools has improved the role of predictive analytics. While traditional analytics looks for significant patterns in data, predictive analytics takes it a step further, analyzing historical and current data to make predictions about future events. For example, professional sports franchises use past data to maximize future ticket revenue. Online marketers have improved customer targeting by extrapolating from past data to predict how customers will react in the future. Sales managers can use historical data to determine the most-optimal future products or territories to assign to each salesperson, and human resources executives can analyze past employee actions to predict which employees are more likely to leave the organization.
Despite the availability of new predictive analytics tools and methods, a 2012 Gartner survey indicated that only 13 percent of organizations make extensive use of predictive analytics. Most companies still focus on traditional analytics to support the business. Furthermore, according to the 2010 IBM Global CFO Study, organizations that run predictive analytics on top of big data will financially outperform their peers by 20 percent or more. With four steps, managers can position themselves to reap the benefits of predictive analytics.
While big data has offered an immense opportunity to expand analytics, the sheer volume of information and number of datatypes can pose challenges. Data analysis techniques will depend on the type of data—structured (from internal applications), semistructured (from e-mail or social media), or unstructured (from audio or video). Poor data quality is a common challenge plaguing the corporate world. Data cleansing and preparation activities are paramount to the successful execution of predictive analysis.
Recent growth in available data and statistical analysis tools has improved the role of predictive analytics—which analyzes historical and current data to make predictions about future events.
The analysis and modeling of predictive analytics require analysts and statisticians with deep expertise in the corporate and operational environment. A McKinsey study found that the demand for analytical talent in the United States could be 50 to 60 percent greater than its projected supply in 2018. Companies can mitigate the risks associated with this talent shortage by identifying potential candidates and providing them with the necessary training.
In fast-paced markets, companies have to be quick and nimble to capitalize on changing trends or risk losing opportunities. If data analysts and statisticians have to spend weeks or months analyzing data and building the right model to determine the path of action, they are likely to miss the opportunity. Powerful new big data technologies and analytical tools are essential for analysts to rationalize and analyze data more effectively.
No amount of analytics helps unless data analysts work in tandem with the business and its objectives. The analysts must coordinate with product and sales teams on a regular basis to obtain guidance on prioritizing the critical decisions. An analytics team can be designed as a cross-functional team composed of data analysts and analysts/managers representing business, finance, and IT. Strong alignment with the business will increase the odds for success and help secure funding for additional analytics efforts.
Suneth Jayawardhane is senior director of insight and customer strategy at Oracle.