ONE GOOD IDEA
The Six Sigma tool to use for high-level product innovation
by K. Sundararajan
The define, measure, analyze, improve and control (DMAIC) and define, measure, analyze, design and verify (DMADV) methods are Six Sigma tools used in new product development when the problem or requirement is fairly well defined.
But when developing products for a new market, innovative ideas are essential. A tool that can be used to spark innovative new product ideas is the ideate, define, design and validate (IDDV) method.
Ideate. In this phase, a group starts the innovation process by thinking outside the box. Brainstorming on a market segment is done to generate new product ideas. Diverging and converging techniques are a good way to create a paradigm shift from which new ideas are developed. For example, a popular soft drink might serve as the benchmark of the fragrance for a new soap. Online Table 1 provides examples of different ways in which new product characteristics could be brainstormed and developed.
Define. In this phase, a team proceeds through the following steps:
1. Identify the product to be developed. Some considerations to keep in mind include:
- What is the need for a product like this?
- Are there missing features in the existing products that this new product could address?
- What are other related product segments?
- What are the benefits of this new product?
The team then uses multi-voting and other tools to select the final idea.
2. The opportunity, goal, development timelines and team members are identified.
3. Process mapping with an action plan is performed.
Design. In this phase, the product design steps are identified. Risk analysis is a good process evaluation method to be used here. As shown in Online Table 2, the inputs needed for the new product or process are evaluated using a risk matrix or failure mode and effects analysis. Based on the score, inputs can be modified to reduce the risk and improve performance.
For example, at the company in which I work, this method was applied to develop a new flavor. The risk matrix for the product is shown in Online Table 3. A cut-off score of 1 x 10-4 was used and ingredient X3 was removed.
In this phase, design of experiments (DoE)1 is used for factor optimization. Each factor is varied at two levels and the output is measured. Factorial experiments are either full (number of experiments = 2 (factors)) or fractional factorials based on Genichi Taguchi’s arrays.2
Response surface measurements are used to further optimize the experimentation. A sequence of DoEs is performed on the factors, and response is improved from point to point until the target is attained.
Validation. This step is necessary to confirm the new product performance. Normally, a pilot scale up is performed as a way to reduce the risk before a full product scale up. In the example of the new flavor development, the flavor was added to a frozen food and consumer-tested. The testing confirmed its superiority over the existing brand.
Six Sigma offers three distinct methods for new product development. Figure 1 shows how you can choose the right method to use based on development time and product novelty required. DMAIC and DMADV should be considered when less development time and product novelty is required.
IDDV requires more development time, but the method can help teams innovate and generate products with the highest levels of uniqueness and originality.
- Anthony Atkinson, Alexander Donev and Randall Tobias, Optimum Experimental Designs, with SAS, Oxford University Press, 2007.
- Ranjit K. Roy, A Primer on the Taguchi Method, Van Nostrand Reinhold, 1990.
K. Sundararajan is the regional quality director for the greater Asia region of International Flavors and Fragrances Inc. in Chennai, India. He has a doctorate in chemistry from the Indian Institute of Technology in Madras. A senior member of ASQ, Sundararajan is an ASQ-certified Six Sigma Black Belt.