Statistical Analysis for Decision Making

Sales, Six Sigma and Statistics: How One Team Used Statistical Analysis to Solve a Sales Challenge and Make Money

Rosa Oppenheim
Contributor: Rosa Oppenheim
Posted: 04/13/2009
It is harder and harder to meet profit and sales projections in this struggling economy. Especially hard if your business is influenced by heavy competition for your products and/or services. We are seeing consumers shift their day-to-day buying decisions, making conscious "trade down" decisions to stretch their dollars. So what can you do to improve your sales performance and manage your growth? More specifically, how can statistics help? The following is a case of an insurance agent who used statistical analysis and Six Sigma process improvement tools to help his firm and increase net income.

First a little background about the insurance agent’s business model and commission structure. It is designed to provide commissions, incentives and bonuses if certain retention objectives are met. In addition to the obvious increase in direct compensation, additional bonus dollars are "triggered" at specified performance levels. For example, the commission would be 6 percent of the premiums paid up through and including 93 percent retention. Once 93 percent retention is exceeded, the commission percentage would go up to possibly 8 percent for the entire retained policy premiums. The payout is further enhanced if the retention of automobile policies meets or exceeds 95 percent. At the 95 percent retention point, an override bonus (lump sum amount) is paid on all policies maintained by the agency. In other words, the office and insurance agent stand to earn significantly more money once these thresholds are reached and exceeded. Seems simple, but we all know that insurance companies do predictive modeling and actuarial tables in the normal course of their business. These thresholds are not arbitrarily set and the insurance firms are pretty confident that achieving these retention rates is no "walk in the park" for the insurance agents.

Using a Six Sigma Process Improvement Approach

Our insurance agent was "hitting the wall" at slightly above 92 percent retention. While it didn’t appear that he had to move the ball far, he had been stuck at the 92.45 percent retention level since 2006. Our insurance agent decided to try a Six Sigma process improvement approach. Working with his office team and a Master Black Belt, they began measuring their current capability. The first step using the Six Sigma process improvement approach was to define the process. The Six Sigma project team got together and after several iterations focused on the key steps of declination. Armed with this information the Six Sigma project team set about measuring the variation. The process focus was condensed to the following steps: (Click on diagram to enlarge.)

The next step of the Six Sigma project was to categorize the reasons for non-renewal. Seventeen different categories were identified and measured. The following table represents the categories identified:

AAuto—Vehicle Stored—Single Item Policy
CAuto—Declination Of Late Payment
DAuto—Driving Record—Insured Or Family
EAuto—Exposures—Including District
FAuto—Insured Dissatisfied With Claim Service
GAuto—Insured Dissatisfied With Coverages Provided
HAuto—Insured Dissatisfied With Premium Amount Charged
IAuto—Purchased Insurance Through Finance Company
JAuto—Rejected Application
KAuto—Sold/Standard Remittance Item In Force Less 60 Days
LAuto—Transfer Out Or New Business Voided To Another Company
MAuto—Vehicle Sold—Single Item Policy
NCancellation U/W Review Period (CURP)
ODebit Balance To Manual Desk For Collection
PNon-Pay—Issued Not Taken Renewal Premium Due
QNon-Pay—Renewal Taken By Other Than Cash—No Subsequent Payment

In addition, during their investigation the Six Sigma project team discovered that some cancellations were not appropriately categorized. Certain cancellation categories are not charged against the insurance agent’s retention targets. Correcting and standardizing the administrative process ensured that cancellations were appropriately categorized.

In the Analyze phase of the Six Sigma project the team found that there was a statistical difference among the categories. Using a Chi Square test the Six Sigma project team determined that there was at least one factor that was statistically different among the categories with a p-value≈0.000. (Click on diagram to enlarge.)

By performing a series of Chi Square tests the Six Sigma project team discovered that four categories/reasons for cancellation were the statistically significant factors creating the variation. Those factors were:
  • Dissatisfaction with premium charged
  • Vehicle sold—In force less than 60 days
  • Vehicle sold—Single item policy
  • Non-pay—Policy Issued, not taken—Renewal premium due (This refers to temporary policies that may be requested on a vehicle for the purposes of sale/title transfer of a vehicle. However, in many cases the sale was not completed, title not conveyed and/or the Client chose to insure the newly acquired vehicle in a different manner after the temporary coverage lapsed)
Armed with this cancellation information, the Six Sigma project team developed processes and strategies to improve retention performance. They launched a pilot that addressed the key issues identified during the analyze phase of the Six Sigma project.

The pilot included some key process changes:
  • Training in and understanding of value options that could be offered to a client to address possible concerns in the cost of the program
  • Rigorous monitoring system of client class and follow-up
  • Changing the work/priority assignments
  • Identification and contact of clients who were approaching the renewal term
  • Proactively providing service and options for the client to consider
The focus on pre-renewal contact for case-by-case evaluation generated significant results. There was also emphasis placed on the "single car" policies and potential cancellations. The changes to the process and specifics of their revised offerings are proprietary but were developed by the Six Sigma project team to address their specific client needs. The big question is, did the changes make a difference? The Six Sigma project team compared performance (using the Chi Square test) of the baseline performance to post-pilot performance and found a statistically verifiable difference, verified by the following Chi Square test result (p-value≈0.000). (Click on diagram to enlarge.)

Results from the Six Sigma Process Improvement Initiative

In the end the pay-off from the Six Sigma project was big. The agency was able to exceed a 95 percent retention level for automobile renewals, which resulted in a net gain to the office of over $160,000. All this from identifying the specific factors causing the variation, tailoring/implementing a solution and measuring performance and progress to sustain the process improvement. Nothing implemented was revolutionary but targeted the focused area for dramatic results.

In these tough economic times an increase like that is welcome, especially when no major costs were added—just an adjustment to the process based on the data, measured consistently and carefully controlled using Six Sigma process improvement and statistical tools to make more money. That is one of the more tangible benefits of the work that we do.
Rosa Oppenheim
Contributor: Rosa Oppenheim
Posted: 04/13/2009


Crowne Plaza Berlin Potsdamer Platz, Berlin, Germany
February 26 - 27, 2018
Millennium and Copthorne Hotels, London, United Kingdom
February 26 - 27, 2018
Horseshoe Bay Resort, Austin, TX, United States
March 25 - 27, 2018