Specifications - Finding Goldilocks: Not too hot, Not too cold, but just right!

Goldilocks can teach us all a few things about specifications. We talk a lot about ensuring we only ship quality to the customer, so that means we must only deliver the highest quality, finest, gold-plated products, and anything else should be scrap, right? Sort of. When thinking about quality it’s important to build in the quality that a customer is willing to pay for.

For instance, while exceeding expectations is great, it would make zero sense to require a 100% CMM inspection of a staple. Sure it would ensure that everything is within a few microns from nominal, but will the customer be willing to pay an extra $100 for a staple with that specification? Of course not.

When approaching specifications, think of it like Goldilocks. Very tight specifications are sometimes necessary, but they are generally more difficult to achieve, and will take additional resources (time, tooling, $$$, etc.) to consistently achieve. Make sure that they are required to delivering the value that customers expect before implementing them onto your products and processes. Because if it’s not required, you’re just being wasteful and it will get incredibly frustrating.

Conversely, overly loose specifications may be easy and cheap to achieve, but will drive more variation and problems into the value stream that will ultimately get delivered to the customer, causing dissatisfaction.

Finding “just right” is critical. And here are a few guiding principles to get there:

• Does the specification satisfy functional requirements, as defined by the customer? If no, then they should be re-assessed. If it satisfies requirements defined by someone other than the customer, you should check with the customer before instituting them into the process, otherwise they may be too high. The monopost 2nd row seats on the Tesla Model X are works of art - beautiful, elegant, and difficult to achieve… but most customers don’t really care about them.
• Is the specification set “because that is the capability of the process”? If so, they should be re-assessed, it’s likely they may be too low. Not all processes have the same capability, and defining specifications based on the process as opposed to the required customer function is a sure way to create unhappy customers. Specifications should come from required function, and processes specified to achieve that function. If no easy process exists that can achieve the function, then either a new process needs to be created that can, or a new method of delivering the same required function to the customer needs to be designed. An existing process doesn’t create the specification, the product definition does.
• Is the specification unknown before you start to build? This may sound silly, but it’s not that crazy - especially with brand new products and new capabilities. It can be difficult to develop a robust set of specifications that you know will result in the right customer value until you’ve been able to start demonstrating that value to a significant number of customers. This is especially true with upstream processes or with new product technologies. In this case, you should aim for as close to nominal as possible, and you should have a specific task force including the Design Engineer, Quality Engineer, Manufacturing Engineer, and the individual responsible for Product, tasked with refining and defining the specification as quickly as possible so manufacturing can scale with confidence.