Big data can already tell you: if you need more staff in the field, what historical repair costs are by year, make and model -- and more.
Big data is everywhere and is becoming more powerful in every aspect of our daily lives. The hunger, and ability, to parse and analyze diverse data sets goes to the highest levels of our government, large corporations and even start-ups, which are finding ways to add immense value. Although we in the claims industry are not involved in international intrigue or attempting to create a surveillance network, we can harness data to effectively influence our decisions. This will ultimately achieve better outcomes by putting specific files in the most skilled hands, streamlining processes and eliminating unnecessary touch-points.
As technology continues to advance exponentially, the auto claims process can be radically improved. For many years, auto insurance claims departments squelched innovative thinking, in favor of tired cost-cutting measures. The departments would look at cycle time, average appraisal cost, repair cost, parts usage and total loss option documentation, but these data points are only a small portion of the picture and don't fully address or reveal the root causes. To get to that, we need to dig a bit deeper.
Today, so much more data can be effectively captured and analyzed that a claims manager could nearly predict outcomes and channel resources for the best results. Need more staffing in the field? What are the historical repair costs of a certain year, make and model with a matching impact location? Where are higher labor rates going to result in more total losses? Which companies allow their policyholders excessive rental on subrogation demands?
Here are a few examples of how auto claims technology and resulting data can be channeled to improve workflows, eliminate unnecessary touch points and create a more efficient process.
- Predictive Dispatching. Imagine putting the right claims into the right hands every time. Through predictive modeling based on a series of data points, the likely repairability of a vehicle can be ascertained to a solid degree from first notice of loss. Combine that with a robust dispatching solution, and the right field representative could be sourced based on factors such as; closest location, past performance, experience levels, historical cycle time results, CSI ratings, estimating software platform, workload volume, current volume for the day and comparisons with other representatives. Algorithms can instantly give the file to the person who is most likely to provide the best outcome. Should the file qualify for a self-service type claim or even trigger a potential total loss, the appropriate resources can be called into action, thus saving money and wasted time.
- Subrogation Demand Analysis. Why do some insurers focus intently on the front end of a claim to ensure accuracy yet will accept subrogation demands from carriers without the same scrutiny? Sure, much of the disconnect may be because of the compartmentalization between various departments within a claims organization, but, with data analysis, subrogation reviews can be parsed, reviewed and stored to track trends. Which companies overpay their insureds for rental when the repair is minor? Do a large number of your files end up in arbitration? If so, what are the triggers? Data can even catch potential fraud if, say, a VIN has been reported multiple times for the same damage.
- Where and when? For quite a while now, most digital cameras have had information embedded in each photograph. With the spread of smartphones and tablets, geocode location data can also be included. What does this mean? If detailed photo information is needed, oftentimes exchangeable image format (EXIF) data can document what brand of camera took the photo, the time it was taken and the exact coordinates. This means a field representative's day can be reconstructed or a re-inspection date can be confirmed to ensure the condition of a vehicle at a specific point in time. While it’s never guaranteed that every phone or camera will provide the exact EXIF information, adjusters will find that, in most situations, a lot of valuable information can be gleaned.
- Claim delay analysis. In auto claims, the standard cycle time for a vehicle damage inspection is 48 hours. Can this always be achieved? If each step of the appraisal process is documented from assignment to completion, any issues and resulting data points along the way can lead to a treasure trove of information. You might discover that an incorrect phone number or vehicle location leads to 25% of all delayed files. This would be a valuable training tool to discuss with adjusters, stressing the importance of documenting accurate information before a dispatch. On the other hand, data may show that a certain percentage of files are delayed because an appraiser is overloaded. This could lead to an immediate focus on better allocating field resources. Capturing and recording specific data points along the way can serve as a tool to detect bottlenecks, inefficiencies and areas for process improvement.
- Location volume analysis. Do you need to increase your field staff? Just think if you could pull up a map at a moment's notice and track volume in a certain location. You could track average repair costs in various zip codes, cities and states and break the information down to an even more granular level. Detailed data can be run over time and compared with newly written policies to predict staffing needs in certain areas before the losses even occur. Think about doing this in real time without needless Excel files and manual processes. This can help in forecasting and budgeting for future years, enabling efficient management and allocation of expenses.
Data, when simplified and made usable, is incredibly powerful. Nary a one of us leaves the house today without a smartphone in his pocket packed with valuable data: phone contacts, mapping directions, family photos, etc. In the claims industry, we must likewise surround ourselves with data, innovating and developing in ways that will let claims leaders manage from quantifiable data instead of basing decisions on emotion and misperceptions.