Is it possible for manufacturers to reduce warranty costs while driving field quality?
The answer is “yes”. Early Warning Systems (EWS) utilize predictive models to alert users across the supply chain about product defects before they turn into costly, hard-to-contain quality issues.
In this post, we will discuss Early Warning Systems – what they are, how they work, and why they are becoming so vital to manufacturing organizations today.
Manufacturers today face heavy competition in an increasingly global war for market share. To succeed, they must focus on four key strategic imperatives – product innovation, speed to market, customer experience, and brand reputation. These strategic imperatives – while vital to driving growth – can put pressure on quality organizations to uncover and quickly address any risks that emerge during any phase of the product lifecycle (see Figure 1 below).
Figure 1: Product Lifecycle
Not only must quality organizations manage internal processes, but they also must understand external dealers’, suppliers’ and service organizations’ processes and systems as well. Engineers and quality analysts – now more than ever – must have the ability to monitor product performance in real-time.
What are Early Warning Systems (EWS)?
Early Warning Systems (EWS) are software systems designed by data scientists who have a deep understanding of warranty data. These experts analyze all relevant warranty and claims data sources and utilize statistical methods to develop “thresholds” (forecasted normal claims rates by product, part and component) and actionable “alarms” (alerts when the claims rate go above the designated threshold by a certain specified deviation).
How do they work?
Early Warning Systems take data from multiple data sources to create an integrated view of Field Quality.
Figure 2: Field Quality Data Sources (Example)
Once the data is aggregated and cleansed, warranty analysts review performance of each product model, part, and year. Often, hundreds of thousands of claim rates are analyzed in order to develop relevant claim “thresholds” and determine how much of a deviation from the threshold should constitute an “alert”.
Early Warning Model – An Example.
A good Early Warning System will identify sudden as well as gradual increases in claims. Consider this example. A manufacturer had a new product that went into production in January 2015, with claim rate data per active unit for every part being entered every two weeks. The Early Warning Model forecasted the “threshold” level of the acceptable claims rate. The claims per active unit crossed above the threshold for the first time in July 2017. Although not a massive spike, it triggered an alert to the quality team to proactively check into defects for that particular part in the field.
A year later, that same part generated nearly triple the threshold number of claims. Had the quality analysts waited until July 2018 to investigate the claims, it would have been too late. Given they investigated a year before, they were able to uncover a faulty component in that particular part, cease production of that part, and replace the faulty component before it led to product failure. The manufacturer also had time to proactively tell customers to bring their products in for servicing in order to replace the part before it led to a costly recall.
Many quality organizations do not use EWS. Instead, they use pareto charts and top 10 lists (e.g. most common repair reasons) to monitor cumulative claims. In these types of analyses, there are no forecasts or predictive models. When cumulative defects reach a certain limit, they trigger “alarms”. This analysis is suboptimal – slow and reactive – because once cumulative defects reach the trigger point (e.g. the July 2018 spike in the example above), it is often too late. Manufacturers can’t stop the production process without significant cost and customer impact. In contrast, early warning systems find issues BEFORE they appear on top 10 lists.
What are the key benefits of Early Warning Systems?
Early Warning Systems enable manufacturers and their quality organizations to benefit from the following:
- An integrated view of warranty, product, manufacturing and customer data
- Shorter issue detection time – from months to weeks
- Improved quality
- Reduced warranty costs
- Higher customer satisfaction due to proactive service
- Reduced risk of damage to brand reputation
If you would like to learn more about Early Warning Systems as well as After, Inc’s other Warranty Analytics solutions, please visit http://afterinc.com/warranty-analytics-solutions/.
You can also click here to download our latest white paper, “Warranty Analytics 2.0 – Addressing the Gaps in Current Warranty Software Solutions”.