In the not-so-distant past, a loan officer shook hands with an applicant, took his measure from across a desk and used his personal judgment to decide whether to risk the bank's funds on a business loan or a mortgage.

Today, most such decisions are based in part on the FICO score, which measures payment history, debt burden, the type of debt a person has, the length of credit history and other factors. The closer your score to 850, the better the credit risk you represent.  

The FICO score is an example of predictive analytics – in fact, the first commercial example of predictive analytics generally known to the public.

Over the past two decades, predictive analytics has been rapidly changing how business decisions are made. The explosive growth in the availability of data and the ability to store and mine data are allowing firms and organizations of all kinds and sizes to begin to use predictive analytics to enhance their business decisions and guide their practice.

The development of a predictive analytics capability within an organization is multifaceted and requires a number of critical linkages. It is a discipline not well suited to dabbling.

• While PA tools are more accessible, the skills to pull off successful deployments are diverse enough to require well-coordinated teams. Lack of a full team profile can lead to inappropriate or insufficient development data, misfires in deployment and negative business impact. While predictive analytics can help you rapidly improve your customer targeting, selection, messaging, sales and business performance, it also has the potential to unravel success in any of these categories with lightning speed. 

• Analysts are central to a solid PA program. The best ones are curious, skeptical, and have the instincts of detectives. They like finding patterns and relationships in data. Recruiting, building and managing a good predictive analytics team requires significant knowledge and skill.

• The capacity to accomplish predictive analytics success also requires a keen understanding of data. You have to understand both the data that your organization or company might have and the relevant data that you might be able to acquire.
  

Organizations or companies just starting out with predictive analytics, and those who have built some analytic capacity but seek to take that to a new level, face a number of challenges ranging from how PA fits into their overall business strategy to the details of recruiting and managing analytic talent. And these challenges and solutions are particular to their specific situation. Success in predictive analytics is a matter of considering all the components and how they fit together.