Analytics turns service repair data into cost savings
With SAS, it takes less than a minute to identify a suspicious claim.
American Honda Motor Co., Inc. uses SAS® to improve warranty claims and forecast usage for parts and services
When a car or truck owner brings a vehicle into an Acura or Honda dealership in the US, there’s more to the visit than a repair or a service check. During each visit, the service technicians generate data on the repairs, including any warranty claims to American Honda Motor Co., Inc., that feed directly into its database. This includes what type of work was performed, what the customer paid, service advisor comments, and many other data points.
Now, multiply this process by dozens of visits a day at over 1,200 dealerships nationwide, and it’s clear – American Honda has big data. It’s up to people like Kendrick Kau, Assistant Manager of American Honda’s Advanced Analytics group, to draw insights from this data and turn it into a useful asset.
Looking backward on a year-by-year basis, we’ve been within 1 percent of where we forecast to be. Kendrick Kau Assistant Manager, Advanced Analytics group Honda
Examining warranty data to make maintenance more efficient
Like any other major automobile distributor, American Honda works with a network of dealerships that perform warrantied repair work on its vehicles. This can be a significant cost for the company, so American Honda uses analytics to make sure that warranty claims are complete and accurate upon submission.
In the case of warranty claims, Kau’s team helps empower dealers to understand the appropriate warranty processes by providing them with useful information via an online report. To support a goal of reducing inappropriate warranty costs, Kau and his team must sift through information on repairs, parts, customers and other details. They chose a visual approach to business intelligence and analytics, powered by SAS, to identify cost reduction opportunities.
To decrease warranty expense, the Advanced Analytics team used SAS Analytics to create a proprietary process to surface suspicious warranty claims for scrutiny on a daily basis to make sure they are in compliance with existing guidelines. The effort to identify and scrutinize claims was once fairly manual, tedious and time-intensive.
“Before SAS, it took one of our staff members one week out of each month to aggregate and report warranty data within Microsoft Excel spreadsheets,” Kau says. “Now, with SAS, we populate those same reports on an easily accessible online dashboard automatically, and we recovered a week of manpower that we could put on other projects.”
Artificial Intelligence Technologies
SAS delivers artificial intelligence through these primary technologies: machine learning, natural language processing, computer vision, forecasting and optimization. Want to learn more about how these technologies play a role in AI?
Identifying suspiscious claims accurately
By applying SAS Analytics to warranty data, the Advanced Analytics group gave the Claims group and field personnel the ability to quickly and accurately identify claims that were incomplete, inaccurate or noncompliant. The results were impressive.
“Initially, it took our examiners over three minutes on average to identify a potentially noncompliant claim, and even then, they were only finding a truly noncompliant claim 35 percent of the time,” Kau says. “Now, with SAS, it takes less than a minute to identify a suspicious claim. And in that time, they are finding a noncompliant claim 76 percent of the time.”
The effort to increase warranty compliance has paid off for American Honda. Through more complete analysis of warranty claims – and more education at the dealerships – American Honda saw a reduction in labor costs for 52 percent of its available labor codes.
Honda – Facts & Figures
Faster claims analysis
Reduced labor costs
Using service data to forecast future needs
The American Honda Advanced Analytics team also uses service and parts data to develop stronger bonds with customers by ensuring dealers have in-demand parts available for customer repairs. Having the right parts available – at the right time – is paramount, so vehicle repairs data feeds directly into American Honda’s marketing and customer retention efforts.
“For the marketing team, we provide strategic insight to help shape their programs that are designed to drive customers to the dealers, and ultimately, keep them loyal to our brand,” Kau says. “The goal of Honda is lifetime owner loyalty. We want our customers to have a good experience, and one of the ways to do that is through exceptional service.”
American Honda uses SAS Forecast Server to assist with business planning to ensure adequate resources are available to meet future demands for services. Using historical information on repair orders and certifications, they developed a time series using years of previous repairs. By combining time series information with sales data, Kau’s team can project where the company’s greatest opportunities are in the years ahead.
“Our goal is to forecast the number of vehicles in operation in order to predict the volume of customers coming into the dealerships,” Kau says. “And that translates to how many parts we should have on hand and helps us to plan staffing to meet customer demands. Looking backward on a year-by-year basis, we’ve been within 1 percent of where we forecast to be. That’s extremely good for a forecast, and I attribute much of that to the abilities of the SAS software.”
Customer feedback that drives the business
Another way American Honda uses analytics is to quickly evaluate customer survey data. Using SAS, the Advanced Analytics team mines survey data to gain insight into how vehicles are being used and identify design changes that are most likely to improve customer satisfaction.
On a weekly basis, the analytics team examines customer survey data. Kau’s team uses SAS to flag emerging trends that may require the attention of design, manufacturing, engineering or other groups. With SAS technology, users can drill down from high-level issues to more specific responses to understand a potential root cause.
“We can look into the data and see what the customers are saying,” Kau says. “And that leads to a number of questions that we can tackle. Is a component designed in the most optimal way? Is it a customer education issue? Is it something that we should address at the manufacturing process? Because of SAS, these are critical questions that we can now identify using our data.”