Using Data Analytics to Maximize Company Performance
Greg Horn | September 14, 2009
Most insurance claims managers keep a close eye on ‘the numbers’—key metrics developed by claims departments used to compare their business performance to their competitors. However reviewing key performance indicators using past performance in order to manage into the future is a bit like driving a car by looking in the rearview mirror.
For that reason, more and more claims managers are examining their own company’s performance just as carefully as the competitors. As proponents of managing physical damage claims by metrics, these managers have developed a number of ways to measure their company’s performance in order to ensure that customer service, claims management and corporate consistency is continually monitored and improved.
Triaging Performance
Claims managers know that if you carefully explore internal (and external) data, this analysis can be used as a valuable tool to gather clues to determine the habits and behaviours of your employees, preferred shops and clients—transforming that rearview mirror into a mechanism that can help you change the road ahead.
In fact, using business analytics to establish a current benchmark for a particular goal can help you uncover your specific training needs and quantify the cost of appraisers and shops that are not performing up to your set standard. For instance, you can actually benchmark the performance of your appraisal staff and shop partners using a particular judgment item, such as the repair of scuffed bumper covers, as long as you have a statistically accurate sample size. This type of detailed assessment and analysis is an incredibly powerful tool that can be used for historical assessment and for future planning.
However, claims managers looking to implement this tool must be sure to exercise caution: Investigate the data sampling to ensure that your sample contains an equivalent distribution of claims. For example, if you include a shop that only deals with heavier hit vehicles, then your sample could be misrepresented, which would skew your results. (You could end up replacing more bumpers than necessary!) Once you have a clear understanding of factors that impact the data collection, you can take these factors into account and start the initial stages of proprietary benchmarking.
By implementing this type of internal data analytics into your operations, you may see a pattern begin to emerge; a pattern that illustrates the differences in performance at various stages in the claims process. This structure also, then, enables your business to analyze what factors are affecting performance.
For example, what if standards were set that could help elevate the lowest performers to meet the company average. Not only would this positively impact your claim department’s bottom line, it would aid your company’s actual dollar value on a year-over-year basis. Another way to examine and set standards is to examine regions or comparable service work.
A Real World Example
An insurance carrier client of Mitchell’s chose to explore this very idea with some impressive results.
Using several data points, we triaged the company’s performance and plotted areas for improvement, settling on remanufactured alloy wheels as the initial area of examination.
Using remanufactured alloy wheels varied dramatically by geography, and there appeared to be a significant dollar amount associated with bringing the lower performing offices up to the average level of performance in the respective geographical areas.
The first performance improvement task was to conduct training. Through analysis it was determined that the training of what could and could not be repaired on alloy wheels could be easily grasped by all levels of service contractors through a one-hour, on-site, hands-on presentation. This cheap method of training was scheduled and conducted during all regularly scheduled staff meeting for a specified time period. After this presentation, data on the service contracts was collected at regular intervals. Initially, re-inspections were conducted on work done to alloy wheels—to ensure the right process and repair had been chosen and conducted. In a short period of time, the metrics on cost and speed of alloy wheel repair and replacement improved.
However, the most significant impact on the use of data analytics was the savings realized by the carrier—savings that topped several million dollars by year’s end. A significant gain just from raising one part of their claims processes performance.
Given the results, the insurer immediately began to perform the same exercise on other areas, and saw similar results.
What was most impressive, however, is that the performance on repair and replacement of alloy wheels did not deteriorate after focus shifted to the new metric—proving that a behaviour change had taken place and, ultimately, resulted in a permanent performance level improvement.
Going from Industry Average to Best in Class
One of the areas I remind clients about when measuring “industry average” is that that performance is, well, average.
For that reason, our clients often choose to benchmark themselves against the best-in-class performers, to measure the pinnacle of performance. However, benchmarking against the best in class must be done carefully.
Research must be conducted to determine who is the best in class and to determine if their business is substantially equivalent to the clients. For example, no standard carrier should be comparing itself to a fleet, as too many variables relating to the type of vehicle repair could cause the data to differ statistically. Fleets, for example, have a much newer and primarily four-door sedan insured book that can be devoid of trucks and SUVs, which would make the sample invalid. Non-standard carriers are also not a good comparison, as they may insure a much older book of business that may have lower representation of OEM products, making potential performance different.
While these metrics are critically important to performance improvement and gaining that edge on your competition, you must exercise great care when choosing a best in class comparison. By careful benchmarking and selecting appropriate areas for improvement, you can begin to use business data analytics to drive money saving performance gains, rather than looking in the rearview mirror.
Greg Horn, vice president of Industry Relations, Mitchell International |
