Data and Details
Getting Sufficient Information to Make It Worth Surfing For.
There’s an oft-quoted saying among statisticians that we are awash in data but starved for information. As editor in chief of Mitchell International’s Industry Trends Report I place a great deal of importance on accurate data, as you can well imagine. I routinely query our data warehouse for latest information that will help me illustrate where the collision industry has been and where it is headed. But there’s more to it than just gathering and interpreting collision data to compile national statistics for use in our magazine. The challenge is to turn the data into useful, actionable information.

Each issue of the Industry Trend Report tracks both the use by parts and dollars per estimate as well as inflationary trends of parts. This macro look at parts utilization and pricing gives an initial insight into parts use across the U.S., illustrating the increases and decreases of alternate and new OEM parts.  

Insurance companies, collision repair shops and car manufacturers are focusing on data to benchmark their findings against their competitors to help them drive improvements in performance. Advances come not only from measuring where they are in comparison to their competitors. With a robust data set, companies also can compare internal performance for various company locations or work units. Detailed data can help a company measure the impact of various initiatives such as a change in part type choice (aftermarket vs. recycled, for example). This can result in a holistic approach where we are able to measure not only the cost of the part but any applicable labor clean- up costs needed as well as any differences in paint and materials costs.   

Going beyond that use, parts providers to the collision market can benefit from the use of targeted data for their market. For example, a salvage yard could examine recycled parts pricing and utilization rates for various popular models to help them better understand potential demand and supply shortages. The more data present for a market area, the better the salvage vehicle decision.

Ongoing data analysis can help a salvage business examine the effect of pricing changes to particular parts in their service market. Equally as important, there are several salvage parts that are not routinely requested from yards. Wiring harnesses and fan shrouds are two examples of parts frequently damaged in collision, but rarely requested by collision repair shops. Data can help analyze and better manage the demand.

Those are just a few examples of data mining that show how data is used today as well as how they could be used with current technology in the collision repair and parts supply business. But that’s just scratching the surface. Data mining is a component of the really hot topic among insurance companies and estimating software companies: predictive analytics. According to a white paper from the American Institute of CPCU, predictive analytics has quickly become an insurance industry best practice and used to target potential clients, to determine more accurate pricing and to identify potentially fraudulent claims.

Predictive analytics is defined as an area of statistical analysis that deals with extracting information from data and using it to predict future trends and behavior patterns. The core of predictive analytics relies on capturing relationships between explanatory variables (also known as independent variables) and the predicted variables from past occurrences, and exploiting them to predict future outcomes. You see this type of analytics when you rent a movie on Netflix or shop online with Amazon or EBay. The vendor gathers data from your past purchases and uses it to predict that you might be interested in some related product or movie. The problem is the accuracy and usability of results will depend greatly on the level of data analysis and the quality of the assumptions. So, for example, if you watched an Indiana Jones movie, Netflix may assume you would like to watch American Graffiti. Why? Because both movies feature Harrison Ford. The two movies have little in common other than one actor. This is a good example of how low volume of data can lead to some bad and potentially costly conclusions. 

Volume is the key to making accurate assumptions. Fortunately, estimating software companies have masses of appraisals with similar collision repairs to draw on. As a result, it would be highly accurate to assume that if the front bumper cover was damaged and the air conditioning compressor was replaced, there is an extremely high probability that the bumper reinforcement should be replaced as well. This could prove to be very important in helping salvage vendors to predict how assemblies should be sold. Assume for example in 97 percent of cases when replacing a three-quarter assembly for a Chevrolet Colorado appraisal that the air mass meter and air box were also damaged. In today’s world you would get the order for the sheet metal but the estimator might assume the air mass meter and air box would need to be purchased from the dealer. If the estimator saw that you had an option in your listing as ”with air mass and air box” or “without,” you would begin to see an increase in estimator awareness of the availability of those parts.

Taking that further, bidding on cars with knowledge gained from predictive analytics could give you very accurate parts demand data on a per vehicle basis, allowing you to better predict your profitability on vehicles you purchased. Equally as important, you would also know when demand for a particular car or assembly begins to wane, allowing you to adjust your behavior accordingly and increase your profit.  

Appraisal data currently has a wealth of potential uses across the entire repair industry, from parts suppliers to repairers and insurers, and the future holds even more promise. Will you be able to capitalize on that promise? The answer will depend on whether you are gathering the right data and whether you are using it to your advantage. Or, as that famous fictional detective and crack analyst Sherlock Holmes once said, “the temptation to form premature theories upon insufficient data is the bane of our profession.”

Greg Horn is Vice President of Industry Relations for Mitchell. Since joining in 2006 he has been a driving force in the company, influencing the design and development of Mitchell’s data‐driven technology, connectivity and information solutions for the Property & Casualty claims and Collision Repair markets.

Click to download the full November/December issue in PDF format

Membership Software By: