Zero Defects is Wrong Approach
There is wisdom in the definition of Six Sigma, which is 3.4 defects per one million opportunities (DPMO), allowing for a 1.5 Sigma shift. But, some companies subscribe to sloganeering such as "Zero Defects". The "Zero Defects" sloganeering is counterproductive, unhelpful, statistically impossible, and completely cost prohibitive.
Statistically, zero defects means a defect level of infinity sigma, which is not possible. What most people mean, is an attitude toward process improvement, but the sloganeering gets in the way.
Are All Defects The Same?
The "Zero Defects" movement has an implicit assumption that all defects are equal. This is not true. In fact, for most firms and products, defects must be identified and prioritized, and attacked and treated from most important to least important. For the defects at the bottom of that prioritized list, it might even make sense to move on and not eliminate or reduce those. The point here is an attitude toward perfection, but fully understanding that perfection is not possible. The attitude and efforts are valuable and the customer will feel and appreciate it. Shareholders will benefit, and the firm will be better for it.
Types of Costs
There are three types of costs that comprise the cost of quality: Appraisal, Preventative, and Failure costs.
Appraisal Costs
The costs in this category includes any and all activities in identifying and assessing for errors or defects in products. For example, testing is an activity that falls in this category; a department that might fall in this category is the Quality Assurance department — this department’s burden would fall under the Appraisal Costs category.
Preventative Costs
The activities that fall under this cost category are training, and any and all activities the encourage prevention or discourages introduction of defects. Establishing processes, procedures, and systems prior to the product being built is typically found in this category. Money spent in this category is money well-spent!
Failure Costs
Failure costs can be both Internal and External. Internal Failure Costs can be money spent to fix defects caught within the firm. For a software firm, money spent on fixing bugs caught prior to shipping a product can fall in the internal failure costs category. For External Failure Costs, these are activities that involve refunds, complaints, call center functions (not outbound sales, but inbound complaints), concessions to the customer for poor service, and warrantees.
Zero Defects and Costs
I present below what I believe to be the relationship between costs and defects:
On the X-axis, we see the costs category (just use a dollar multiplier for the X tick marks). On the Y-axis, we see defects, by count. So, we see that as defects approach zero, costs increases exponentially and hovers asymptotically on the x-axis, and never reaches zero.
Footnote: The costs to a firm where there is no effort to identify and reduce errors or defects can also be exponential. For example, imagine a firm where there was no inspection, appraisal, or prevention of faulty, defective, or harmful medical devices or drugs. The external failure costs alone could bring the firm to bankruptcy. I recognize this fact, but wanted to make a point above regarding zero defects.
Why is the above Graph True?
As defects are identified and eliminated, there will be theoretically few defects. But, this means that identifying defects will require more effort and will become more and more difficult, thus increasing the costs of this activity, along with the subsequent costs to fix the defects identified: The costs to inspect and test increases as there are fewer and fewer defects.
Conclusion
Sloganeering doesn’t help, especially if the slogan makes no sense. "Zero Defects" as a mantra has a nice cumbaya ring to it, but doesn’t really help or motivate a crew to do better. Moreover, "Zero Defects" is statistically impossible as well as cost prohibitive.
Regarding defects: not all defects are equal. It is important to identify the defects that impact the customer, prioritize those, then respond in a prudent way in improving the product or process. Going after all defects is not prudent.
The key takeaway here is the following: strive to be better everyday; strive to make the customer happy. The Firm’s efforts to make the customer happy will be felt, a culture of improvement will be created, and the firm and shareholders will benefit from it.
WP Cumulus Flash tag cloud by Roy Tanck and Luke Morton requires Flash Player 9 or better.
If you enjoyed this post, please consider to leave a comment or subscribe to the feed and get future articles delivered to your feed reader.
Comments
I blogged the following recently on this topic (http://tinyurl.com/2df4tl):
Fundamentally, it needs to be appreciated that there is a difference between reducing defects and eliminating them. The cost of trying to support zero defects goes up exponentially as you get closer to zero. Deming used the red bead experiment to highlight that if a process has a failure rate built into it from the beginning, worker competence is not the issue. It really is all about management.
Rob
I think there’s a difference between “zero tolerance” and a goal of “zero defects.” Zero tolerance is a dumb policy if you’re going to be overly punitive and point fingers (or fire people) every time a mistake is made. We’re human. The Toyota Production System recognizes this.
The goal of zero defects can be achieved through Error Proofing devices or methods. The most clever error proofing methods are inexpensive — think things like a template bar that prevents a too-tall vehicle from driving into a parking garage when it wouldn’t have fit. Error Proofing methods are more cost effective than traditional defect prevention methods, such as inspection or telling employees to “be careful or you’re fired.”
@Mark,
Poka-Yoke is absoutely the way to go for newly-created processes, or as an solution for already-created processes but are producing defects. For mature firms, however, there exists processes, methods, tribal knowledge, and culture, that need help. Yes, prevent errors through mistake-proofing; for processes that have already been engineered but are producing defective products — that was really the point of this article.
Poka-Yoke can also be applied to existing processes, and not just in manufacturing. You can apply poka-yoke to business processes, as well.
Peter’s graph is precisely why the founders of the Toyota Production System emphasized built-in quality (jidoka). Arguably more important than error-proofing (pokayoke), built-in quality is the antidote to the futility of “quality by inspection”.
In the Lean community, I personally don’t feel the proper emphasis being given to built-in quality, and I’m not sure why.
The Taguchi Loss Function demonstrates that quality cannot be inspected in, but it can be controlled through robust design. Likewise, proper equipment design can catch quality defects and stop the process before they are passed on to the customer (or the next process).
Mark is absolutely right that you can error-proof any business process, but I’m more interested in why companies don’t emphasize DFx in process design like they do for product design.
Building quality into business processes is about the fundamentals — robust design and visual management (what makes any generic business process stop when quality is unacceptable).
More about Taguchi Quality Methods
http://en.wikipedia.org/wiki/Taguchi_methods
http://www.sixsigmafirst.com/intro2_taguchi1.htm
[...] One of the posts there is rather interesting. Pete Abila of shmula wrote about Zero Defects and it being statistically impossible. He says: The “Zero Defects” movement has an implicit assumption that all defects are equal. This is not true. In fact, for most firms and products, defects must be identified and prioritized, and attacked and treated from most important to least important. For the defects at the bottom of that prioritized list, it might even make sense to move on and not eliminate or reduce those. The point here is an attitude toward perfection, but fully understanding that perfection is not possible. The attitude and efforts are valuable and the customer will feel and appreciate it. Shareholders will benefit, and the firm will be better for it. [...]
[...] There has been a lot of buzz surrounding the attainment of zero defects recently. I’ve written about this topic before but this post summaries my own position nicely: Effectively zero defects is not really achievable in most cases. Defects are largely a matter of definition. As performance improves expectations will often rise. When you eliminate anything you would have called a defect years ago, standards are higher and things that would not have been called defects are no longer acceptable. At some point the system process advances to such a level where zero defects is possible in some cases but in many (say medical care, air transportation, education, computer software, restaurants, government, management consulting, civil engineering, legal services…) I really think it is basically impossible. [...]
[...] than fixing problems, but only for 95% of the problems. In fact, it has been noted before that zero defects is the wrong approach, because preventing those last few problems is far too expensive. Which means that we have to allow [...]
Your post completely ignores the fact that if you could achieve Zero Defects the process would be not only faster, but immensely cheaper!
Also, in your cost graph you do not account (at all) for the cost of leaving the defects in the product. That cost would be much higher when the number of defects is higher (assuming the same distribution between trivial/critical defects)
Here’s my full argument as to why aiming for Zero Defects is not only good, it’s the only way to survive long term: http://bit.ly/15lkle
Leave a comment
Additional comments powered by BackType
I had the pleasure of working at a company that tried to enforce a ‘zero tolerance’ stance on software and editorial defects.
I certainly will never institute such a policy at my current employer for our developers. To work towards approaching zero (regardless of whether that number is 10, 100, 1000 or whatever is right for the resources you have at hand and the value it delivers to the business) is a much better approach in practice.
It wont surprise you that my previous employer is no longer operating…