Monday, 25 July 2011 16:47

The Devil in the Details—Measuring Metrics Properly

The Q2 2011 Mitchell Industry Trends Report (ITR) contained a feature I wrote about how repairable estimates can best be executed, “Are You Measuring Your Auto Insurance Claims Metrics Properly?” I argued (persuasively, I hope) that defined metrics and a clear plan can help make your business great, achieving measurably better auto insurance claims processing outcomes in the process.

To view the full text of this article with photos please click HERE.

The key is to understand who your competitors are and if you can replicate their performance. Only then can you have an achievable performance improvement goal that allows you to focus on the person, part and process that can be improved.

Determining what is great estimating performance, and whether this great performance can be replicated, requires a large sample size to allow for a statistically significant evaluation. In addition, the ability to drill down and segregate vehicles by type, age and country of origin as well as factoring in the geography of the repair estimates are all necessary in order to truly have an accurate comparison.

Comparing your performance against the industry average by identifying the best performers isn’t a simple, mindless process, and nor is analyzing the data and setting performance benchmarks. Adding to the complexity is the constantly changing mix of vehicles coming through the repair process.

As an example of how specific metrics need to be in order to be meaningful, consider this—there can be more than an hour difference in labor time in Removing and Replacing a used bumper assembly vs. overhauling an existing damaged bumper system. That hour difference can be important in an area with extremely high labor rates.


To measure metrics properly, looking at the distribution of the claims by inspection type (sometimes called appraisal channel) can help you let the data help your business. It is quite revealing in terms of distribution of claims severity. Removing the lower dollar estimates alters the overall average paid severity (not sending independent appraisers out on small hits and relying on claimants to fax in estimates for example), reducing repair vs. replace, paint hours metrics and part use when comparing estimates to appraisal channels that routinely see those smaller hits.

Similarly, Direct Repair average paid severity is actually a measure of the completed repair orders and not the average of an appraisal written by a staff appraiser. Why is that? Because some claimants will make a claim and not repair the vehicle, so any hidden damage or price changes would not be reflected in these appraisal totals.

Again, no easy task to ensure the data is as close as possible for comparison, and next up is measurement scale: do you want to measure your company performance against the average or the best in class? Both are advisable; measuring merely against the industry average only provides a comparison to the middle, and not the best that can be achieved.

Rather, you need to find where the best performers are in your given data subsets, those in the 75th percentile and above. Establishing a best- in-class goal that is achievable because it considers the types of businesses (standard vs. non-standard insurer, vehicle mix, etc.) you are indicating are your closest competitors along with other factors, such as geography, is a worthy goal – one that will help you meet your particular performance goals.

All the data means something. The key is identifying what the data means to your business performance. And this brings us to the importance of understanding what to measure. Many industry measurements and metrics are somewhat flawed, and need to be revisited, or refined with a second measurement, in the name of proper measurement and dependable metrics. In addition, many best-in-class metrics are a moving target. There are multiple factors in flux that keep the definition of best performance ever-changing: Americans are moving towards smaller and more fuel-efficient cars, so the repair vs. replacement of panels performance metric will change. Smaller cars, thinner gauge metals, changing metallurgy, and the low cost of alternate parts will reduce repair opportunities.

On the topic of changing metrics, currently we see OEMs and their price-matching programs having an interesting impact on market share assessments. OEMs’ price-matching programs help them sell more parts. But lowered OEM parts prices make it look as though OEMs are losing market share to aftermarket and used parts, at a faster rate than measuring the number of OEM parts specified on the average repair. In fact, today the percentage of alternate parts use on estimates is overstated when using the industry standard metric of ‘percentage of parts dollars’. When measuring the number of parts by part type, we see that used parts have remained fairly steady at a little under 1 part per estimate, and aftermarket parts at bit over 1 part per estimate.  Looking at parts use in this way gives us another important measure of parts performance and provides better visibility into the performance metrics that matter to your business.

The industry needs more complete picture of parts use. I suggest keeping the existing metric and industry standard and then for true insight, add the number of parts on the estimate by parts type. This approach does go against the prevailing wisdom, but understanding the real metric here is important—because by supplying the majority of parts on a Repair Order, the OEMs are the pivotal parties of cycle time, and shops need to maintain good communication with all of their parts vendors and understand what is in stock and when the delivery will be made.

Starting in the third quarter of 2011, the next Mitchell Industry Trends Report will include this new metric, in a bid to help tame the devil in the details and help the industry measure metrics properly.

Statements and opinions expressed in this article are solely those of the author.  They are not offered as and do not constitute legal advice or opinion of Mitchell International, Inc.

Read 3774 times