If you want to reduce warranty claims for your products, there are very few roadmaps available for this kind of effort. I’ve personally been involved in three major multi-year efforts of this type with large corporations, and I’ve learned more than a bit about what does and doesn’t work.
Let’s start by considering three basic ideas you should be aware of if you are asked to lead or participate in a significant attempt to reduce warranty for your company’s product lines. Here they are:
A—YOU WON’T BE POPULAR
If you are asked to lead or support major warranty reduction activities, don’t expect to be loved for doing so, at least not for some time. In fact, it may take years before your work is appreciated, even by the executives who asked you to take this on.
Why? To start, highlighting warranty claims and expenses will indirectly criticize a large number of people and organizations in the business. No matter how you how it’s presented, focusing on warranty focuses on failure. Few will respond positively. Many will be out-and-out hostile. Be prepared. Don’t respond with anger, either. That will make things worse.
You are almost certain to discover that everyone, and I mean everyone, who needs to act to resolve a problem will question your every move. They’ll nit-pick every presentation, debate every metric, and they’ll even question your motives. This will include field personnel who work with customers daily, because they’ll vent their dissatisfaction with the situation to you as the representative of “headquarters” or whatever the central location of the business is called.
B—YOU NEED DATA, DATA, AND MORE DATA
A company suffering from excess warranty claims and costs needs data to make anything other than educated guesses about the scope and nature of the issues they face. However, even if a warranty data system exists, you are all but certain to learn several hard lessons about the data that has been collected—or about new data collection methods you initiate.
You will probably learn that the information in any warranty tracking database is riddled with errors. Some are trivial; many are systematic. And more than a bit of the information will be downright misleading. If you can get to the point where you have 70-80% confidence in the accuracy of the data, you are doing very well indeed.
You’ll also need to take a very hard look at the metrics you extract from warranty data. To paraphrase Dr. W. E. Deming, there is no such thing as a warranty metric until you develop an operational definition for each and every metric. And, you will quickly learn that one simple number—something that managers and executives will want to see—will never tell you enough to understand what’s really going on or whether you are making any progress.
You’ll probably have to settle on about a half-dozen metrics to have any real hope of being able to see the big picture, or even a small picture for that matter.
You will also find that obtaining and analyzing returned goods that have failed will rarely tell you the whole story. They may also lead to confusion and debate about any analyses that are done. That doesn’t mean you shouldn’t try to understand specific failures, but, in the absence of broad-based metrics, field return analyses aren’t as revealing as you might think.
C—IT WILL TAKE LONGER THAN ANY EXECUTIVE IS EXPECTING
Even when you succeed, your results won’t bring instant financial payback. While a major warranty problem can be created in a very short time, it nearly always takes years to see the full impact of corrective action.
When a new or altered product enters the marketplace, warranty issues are unlikely to appear immediately. The worst problems can’t be identified with any confidence for one, two, or even three years after a new product is sold to end customers. Then, it will take more time to understand the cause-and-effect relationships that led to an issue (and there are many pitfalls in doing that).
A change in design and/or production will be needed. Picking and validating the right “fix” takes time and may lead to extensive debate and/or handwringing. You can easily make things worse if you botch this phase of the cycle.
At every point in the supply chain, existing inventory will have to be dealt with, and no one will want to scrap anything unless it’s hopelessly flawed. There will be resistance to rework, too. None of these actions are in anyone’s budget or have been included in their annual performance plans, so expect stiff resistance.
Finally, new-and-improved products have to enter the production stream and be sold. Then it will take the same amount of time needed to see the problem in warranty results to confirm that the corrective action was effective. So, if it took 3 years to see the issue clearly, it would take another 3-4 years to see if you actually did anything positive.
At the same time, everything produced in the time consumed to see the issue clearly, formulate a corrective action, and get the revised product into the marketplace will continue to add to the warranty mountain you are climbing.
The finance staff won’t be able to book any cost reduction results for a long time. A discounted cash flow based on the costs associated with diagnosis and correction will be unlikely to show a rate of return that excites anyone. That will make financial justification for changes difficult. This will also test the patience of those who chartered your activities.
Making progress is slow, difficult and, unless you like irritating people, frustrating. I’ll expand on how you can do better in Livonia Technical Services articles–coming soon.