Reducing Claim Costs for Harley- Davidson
Reducing Claim Costs for Harley-Davidson by over $800,000 through Best-in-Class Warranty Analytics
Since 1903, Harley-Davidson has defined motorcycle culture with an expanding range of leading-edge, distinctive and customizable motorcycles in addition to riding experiences and exceptional motorcycle accessories, riding gear and apparel.
Harley-Davidson Financial Services provides financing, insurance and other programs to help get Harley-Davidson riders on the road.
-
$5.7billion annual evenue (2018)
-
100global markets
-
5,600employees worldwide
-
Business Challenge
Harley-Davidson, like most large vehicle manufacturers, is focusing on eliminating unnecessary and excessive Warranty Program costs. In 2017, Harley’s Global Warranty Manager was searching for a partner to help identify claims cost reduction opportunities and provide tangible, tactical recommendations. Given After, Inc.’s reputation as a leading provider of Warranty Analytics Services to the world’s largest manufacturers, Harley requested our support.
-
Solution
After, Inc. proposed a Proof of Concept (POC) to uncover potential over-repair and/or fraud. Given After’s experience with dealer claims analysis (for customers like Ford, Jaguar Land Rover, and others), we predicted there was likely some low-hanging fruit in the way of dealer over-repair/fraud that could be identified through statistical analysis. Once the After Team looked at Harley-Davidson’s Warranty Claims process and learned that its Warranty Team was auditing claims every quarter and rarely finding significant issues, we felt even more confident in our approach.
The methodology for Harley’s Proof of Concept (POC) involved After, Inc.’s best-in-class five-step approach:
1. Gather, Integrate and Evaluate the Data
After, Inc. took a year’s worth of dealer claims data and integrated it into a single database for analysis.2. Establish Benchmarks
Next, we segmented the data to establish benchmarks for each type of repair based on parts and labor costs as well as repair efficiency. This involved significant modeling to make it easier to compare dealers in different geographic areas or with different product/service mixes.3. Identify Deviations
After, Inc. looked at the data nationally, as well as broken down by region, model/sub-model, repair type, dealer/servicer, and warranty quarter. Through each type of analysis, we identified dealers that had significant deviations from the average benchmarks (“outliers”).4. Conduct Text-Based Analysis
Text-based analysis is an important step in our approach that helps reveal potential inconsistencies between technician comments and repairs performed. In Harley-Davidson’s case, some dealers repeated technician’s exact comments – word for word – multiple times which was a red flag for over-repair raised concern for fraud or over-repair.5. Execute Thoughtful Communication Plan
As a next step, we outlined how Harley could devise policy changes that would ensure that all dealers “self-corrected” and establish consequences for ignoring the policies. -
Results
Agreed to a multi-year Warranty Analytics contract.
After a successful POC in late 2018, Harley agreed to a multi-year contact with After, Inc. whereby After would help them reduce dealer claims costs, minimize over-repair, and institute policy changes to support long-term fraud prevention. We would also support quarterly dealer reviews and other cost control initiatives.
Exceeded client’s ROI expectations.
The partnership allowed Harley to have greater success with dealer audits, prevent dealers from over-repair on certain jobs and make changes to their policies. As a result, in the first full year of the contract, After, Inc. was able to help Harley-Davidson reduce its claim expense by over $800,000 or 3.6%.
Next up: Implement real-time claims scoring.
Next step for our partnership is to implement real-time claims scoring. After’s machine learning models will identify suspicious claims during the adjudication process – before payments are made instead of after the fact.