Imagine you are the swimming coach for the US Olympic team. You have a pool full of great swimmers with endless potential, and you want to come home with as many gold medals as possible. When you are deciding which participants to put on the medley relay, where four strokes are represented and a variety of talents are needed, you wouldn’t just randomly choose four swimmers. Even picking the four best swimmers might not give you the best chance of winning the gold. You would identify your best swimmers and which swimmer performs the best at each particular stroke. You might also consider how well those four swimmers work as a team, how many other races each swimmer is participating in, and which swimmers hold up best under pressure. The team who comes home with the most gold medals not only knows who their best swimmers are but optimally allocates swimmers for each race.

It is not just coaches who use optimization to get optimal results. The field of Operations Research, where mathematical models and advanced analytics are used to identify the most optimal decision, is employed across many fields.  By using advanced analytics, the sales force at NBC reduced rework by over 80% and increased revenue by over $200 million between 1996 and 2000 [1]. Similarly, Motorola changed the way it conducted negotiations with suppliers to an analytics-driven platform with online negotiations and scenario-based optimization analysis in the early 2000s, resulting in a savings of more than $600 million [2].

Accounts Receivable Management is an area prime for operations research. The uses of operations research and advanced analytics in accounts receivable management are numerous, from reducing mail costs to using a score, such as the NLP Logix predictive modeling score, to construct your overall dialer strategy. Today, we will focus on one specific strategy: using advanced analytics to optimize your staffing strategy.

Like the swimming coach, the first step in optimizing your staffing strategy is to figure out what “races” you’ll be in. If this is your first foray into advanced analytics, you might start out with a simple segmentation method using customer location or business unit. Otherwise, you might use a more sophisticated machine learning method like clustering to identify customers with similar behaviors (more details on that will be in a later blog). Once you’ve determined how you’re going to segment accounts, you need to identify what proportion of your business each segment represents and what team size is needed for that segment.

You next need to identify your best people. Using advanced analytics, NLP Logix can help you to develop metrics that measure your agents’ successes. By combining your agency’s historical dialer, financial, and time-clock data, we can identify agents who maximize conversions or liquidity while minimizing lost opportunity and non-productive time. These same characteristics are then recalculated for each agent based on their historical performance with each of your customer segments. In the same way that the coach isn’t just interested in which swimmer is best overall, but rather which is best at each stroke, we’re interested in which agents will perform the best for each customer segment.

Once we’ve identified how each agent performs for each customer segment, we can then use operations research techniques to identify the optimal allocation of employees to each team. At this point, we can also incorporate other variables that might affect employee performance. By understanding how well the employee performs based on factors like time of day or shift length, we can optimize the staffing method that will maintain the coverage you need while maximizing collections.  As you start collecting data on your newly allocated teams, we can even begin to better understand the customer segments your teams are working. This can help you understand if one group prefers a short talk-off while another might respond better to a long talk-off, or if one group only responds to settlement offers while another prefers to pay in full.

When these operations research and analytics methods are combined with the NLP Logix predictive modeling score, goal setting for employees is vastly simplified. Because the expected value can be calculated for each incoming account, goals can be established that push the employees to perform while accounting for the amount they should actually be able to collect from the particular debt they’re working!

If you’re ready to dive in to creating a gold medal team allocation, ask NLP Logix about data-driven decisions.



Next Article: Judge Dismisses FDCPA Credit Bureau Reporting Case ...

For more from NLP Logix, visit their blog