Previously, we had discussed quality’s role in New Product Development (NPD) and outlined a few ways in which quality can meaningfully contribute to NPD success. We left a cliff-hanger in that post, which will be addressed here. How does Machine Learning intersect and connect quality and NPD?
Old Quality Answers to Old NPD Questions
My colleague, Andrew Hughes has recently worked on a four-square framework to assess analytics by types of questions they answer. It’s a great way to understand the value of Machine Learning, and is also a good framework to understand and guide your company’s analytics capabilities. Companies don’t just jump to the front of the line by deploying Machine Learning alone.
Quality leaders should start from the bottom left corner. How well do we answer Old Questions such as “what is the Voice of Customer (VOC)” and “what should requirements and specifications be?” One Old Answer is to report quantities and types of Non-Conformance and Complaints that provide guidance on past conformance failures. This information is used to design out these issues and therefore is an elegant preventive action.
A better Old Answer can be provided by associating NCs and Complaints to Specifications, Parts, Processes, Suppliers, etc. to identify problem areas. This better informs the new product team by identifying where additional process capabilities may be needed, compare part performance, grade supplier performance, and possibly identify where specifications may have been set improperly. This requires more sophistication. While it is theoretically possible without IT investment, a pragmatic way to address this is through an Enterprise Quality Management Software (EQMS) connected to Product Lifecycle Management (PLM), Manufacturing Operations Management (MOM), Customer Relationship Management (CRM), etc.
New Quality Answers to Old NPD Questions
Our better Old Answer provides good insight into Specifications and some insight into VOC. It provides well-informed general patterns. Useful information, but things change, and not all parts of a decision surface follow the same pattern. Wouldn’t it be much better to know the conditions in the moment? What if we could “fail fast, fail often” and quickly innovate by understanding customers’ reactions in a more meaningful and real-time fashion?
Also, if we want market success, we need to take inputs from the broader market. Therefore, looking just at Customer Complaints from the existing customer base would be insufficient. What about other demographics or geographies? What do those prospective customers like and complain about with competitive products?
Using the Industrial Internet of Things (IIoT) and Machine Learning, we can provide fresh perspectives on these old NPD questions.
What Should Quality Leaders Do?
23% of companies (N=252) are investing in the IIoT to drive Quality Improvement today, and quality improvement starts even before the concept at the VOC. This is one of the top IIoT use cases, which means that Quality has a seat at the IIoT table. Leaders should:
- Recognize that Quality Improvement is a top IIoT use case and that IIoT is a strategic initiative they must leverage to drive change
- Use this opportunity to shift corporate mindset away from “quality is a department” / “quality ensures compliance” to “quality is a cross-functional responsibility led by the quality department” / “quality enables success and value”
- Ensure that quality can provide meaningful Old Quality Answers to Old NPD Questions through a centralized and connected EQMS
- Leverage IIoT to enable meaningful participation in VOC by providing New Answers to Old Questions
The beginning of the NPD process is and risky and critical phase in the product lifecycle. Through more meaningful quality engagement in VOC, by investing to provide good Old Answers to Old Questions, then growing to provide New Answers to Old Questions, quality leaders can reduce risk and deliver success and value.