Creating Continuous Innovation - ASQ

Creating Continuous Innovation


Download the Article (PDF, 53 KB)

Mark P. Finster, University Of Wisconsin

Cole conceptualizes continuous innovation (CI) by contrasting continuous improvement, discontinuous innovation (DI), incremental innovation, exploitation, and exploration. He notes that these research-oriented contrasts are synergies in practice, best managed holistically as parts of an organization’s approach to improvement. Cole suggests that DI in new product development may become more continuous by “smoothing” innovation, introducing an iterative sequence of incremental steps, each informing subsequent steps about the nature and evolution of the marketplace, thereby shortening product development cycle-times and reducing the costs and risks involved. He emphasizes that CI involves early stages of the product development process with a strong focus on the customer, and he describes approaches that transform DI practices to CI.

This paper broadens the concept of CI beyond new product development by linking four research streams—creativity, sociotechnical system design, design of experiments, and offensive quality—to CI. This paper also discusses future research issues related to CI and addresses three contrasts that add depth to CI’s conceptual foundation: CI with continual improvement, creativity with innovation, and defensive versus offensive quality. Note that the use of these terms is philosophical. The terms are not defined operationally to categorize a change event as either meeting or failing to meet the definition.


W. Edwards Deming described improvement as a continual part of everyone’s job, everywhere in an organization, from top to bottom, horizontally across all functions, throughout the entire supply and delivery chains, in all dimensions, including cost, product quality, response time, employee morale, and the external environment (Deming 1986, 49–52). In contrast with CI, continual improvement encompasses all forms of improvement, including (1) elimination of defects (anything internal or external customers do not like); (2) improvements that incrementally add value to customers; and (3) larger-scale, longer-term innovations, often strategic in nature, that dramatically redefine systems, processes, and outputs (Imai 1986). In Cole’s vernacular, (1) is primarily continuous improvement, (2) is mostly incremental improvement and CI, while (3) is discontinuous innovation. Juran’s trilogy demonstrates how these aspects of improvement fit together (Juran 1992).

For example, suppose a mail-order retail store like L. L. Bean innovates by developing a Web-based storefront that enhances its sales channels. Implementation requires integrated back-shop innovations in the operations and information systems of its entire supply chain, including hardware and software changes to accommodate newly established electronic links and the expected increase in sales volume. Customer response time is already tracked and controlled as a key performance quality indicator. However, the targeted response time must dramatically decrease in order to remain competitive with the more stringent customer expectations of that Web channel. Response time, when viewed operationally in the pick department, relates critically to another key customer requirement—accuracy. The pick line requires incremental improvements to shorten lead times, and accuracy is controlled by electronically weighing a filled order prior to its shipment. Large deviations in weight signal an out-of-control order whose accuracy must be corrected. Since product weight is a control variable that translates to customer value (accuracy and response time), all supply chain producers must manage product weight as a key quality control variable while developing innovative operational and IT solutions. Based on volume projections and lead time data, those suppliers with long lead times are targeted for collaborative innovation, while suppliers with short lead times are asked to improve incrementally.

This example illustrates several points. First, the various types of improvement synergize. Second, all innovation involves incremental and continuous improvement. Third, innovation requires more system thinking than incremental improvement. Fourth, customer qualities and control variables do not necessarily change during innovation, although their targeted values may change dramatically. Fifth, Web channel development is another example where CI can be a valuable approach for improvement.


Creativity researchers contrast creativity with innovation, describing creativity as the process for developing original and valuable ideas. The five major research journals on creativity consistently define creative ideas as those that are original and valuable. In this sense the concept of creativity sits in the eye of the beholder, much like quality. Creativity scholars study the benefits of creativity (which include return on investment, market share, customer and employee satisfaction, retention and loyalty, employee self-actualization, and contribution to change efforts); methods for assessing creativity; the mental and process characteristics of creative problem-solving approaches; the environments that support creativity; and the personal characteristics that lead to creative output (Isaksen et al. 1994). Scholars of innovation study the processes by which these creative ideas develop and diffuse into organizations, industries, and countries, as well as the mechanisms and processes by which creative ideas are adopted and implemented (Rogers 1995). According to Rodgers (1995), key factors include leadership characteristics (like attitude and support toward change) and organizational factors like degree of centralization, complexity (for example, job diversity and professionalism), formalization and rigidity (which inhibit diffusion), interconnectedness (the informal social network that enhances diffusion), and organizational slack (the degree to which resources are available). Both research streams support and contribute to Cole’s concept of CI.

For example, the creativity literature identifies two approaches to creativity. One approach identifies those who are most likely to be creative and then employs those candidates in organizational positions requiring creative output, positions like research and development, product conceptualization, and advertising. The second approach to creativity discusses processes, environment, mechanisms, and tools for raising the creative capacity of everyone in an organization. This approach is inclusive, much like the approach to continuous improvement and Cole’s approach to CI, seeking to involve everyone rather than keep innovation in the hands of a few experts. Creativity research also supports evolutionary stages for building creative capacity, beginning with employee participation (behavior), evolving to involvement (head and heart), and culminating in empowerment (capable ownership).


While CI is introduced as a product development strategy, it can extend to innovative system design. Sociotechnical system (STS) researchers advocate holistic approaches to innovative system design that effectively integrate the technical, social, and business systems of an organization (Taylor and Felton 1993; Pasmore 1988). STS scholars develop the philosophy, processes, and tools of effective design for system innovation. Their research findings support the need for extensive employee involvement in creation and innovation processes (Table 1) (Huber and Brown 1991, 142). The last principle emphasizes the continuous nature of system design and supports Cole’s premise that innovation is a continuous process.


The third contrast that develops CI compares defensive quality with offensive quality (Kano 1987; Kano et al. 1996). Defensive quality creates value by focusing on the elimination of customer dissatisfaction, caused primarily by failing to meet customer expectations. Defensive quality seeks stability by eliminating and reducing non–value-added activities that lead to added costs and customer dissatisfaction. At the core of defensive quality is continuous improvement.

For example, defensive quality is the approach used by six sigma programs. After training and infrastructure development, six sigma programs involve all employees in developing targets and metrics that operationally define their customers’ likes and dislikes. Six sigma programs then proceed to identify and continuously improve the processes critical to these customer targets so as to reduce and eliminate variation, thereby reducing customer dissatisfaction.

However, defensive strategies, like most six sigma programs, fail to address product and service innovation. General Electric closed this gap by developing Design for Six Sigma. DFSS integrates approaches for innovation into product design processes (GEMS 1999). DFSS involves greater emphasis on collection and analysis of market information, precise translation into the value characteristics of the product critical to quality, quality function deployment, rapid prototyping, development of transfer functions for rapid deployment to operations, and concurrent engineering. Like CI, these technologies seek to rapidly deliver value to customers in innovative ways.

DFSS is an example of offensive quality. Offensive quality approaches produce customer value by creating what customers do like, in contrast with defensive quality approaches that eliminate what customers do not like (variation from target). While defensive quality makes existing processes and products error free, offensive quality involves creativity and innovation to introduce new products and services, and bundles of new product features (Kano 1987; Kano et al. 1996). CI, as defined by Cole, is primarily a strategy for offensive quality. Note that many quality researchers and practitioners are actively involved in offensive quality, as exemplified by the 12 annual joint conferences between the Theory of Inventive Problem Solving (TRIZ) creativity experts and the QFD Symposium held each year in Novi, Michigan.


Cole emphasizes that CI “occurs in the early stages of product development” and “uses the customer as a driving force for the learning process.” Successful offensive quality strategies, such as CI, require thorough investigation into customer behavior, probing, and exploring to discover opportunities for adding new value. Unlike defensive quality, which asks customers to articulate what they like and dislike in current products, offensive quality often involves areas of value that customers themselves do not understand, cannot articulate, or are unable to recognize (Leonard and Swap 1999, chapter 3).

For highly innovative products, organizations must be wary of preferences expressed by customers when they view prototypes. During market tests of prototypes, customers overwhelmingly rejected push-button phones and Sony’s popular Walkman devices. In the early 1990s, Compaq bet millions on PC network servers even though customers said they would never abandon their mainframes. Compaq persisted and by 1994 sold $1.8 billion worth of these servers. New Coke scored higher than Classic Coke in all consumer tests but was soundly rejected in the marketplace. Spaulding, Mizuno, and Rawlings devised a baseball glove that “pumped-up,” mimicking the pump sneaker, to create a tight fit. Consumer focus groups loved it, but nobody bought it. Fax machines, VCRs, FedEx, CNN, and telephones are just a few innovations consumers rejected in focus groups with prototypes (Justin 1995).

Therefore market researchers engaged in CI must probe and explore beyond traditional customer preference inquiries to develop approaches for creating offensive quality. Traditional market research provides prototypes as stimuli, solicits customer preferences as response data, and then makes decisions regarding target markets, product family mix, product configurations, and bundles of features based on analysis of this preference data. The more innovative the product (discontinuous innovation), the more important it is to understand the reasoning behind customer preferences (cause analysis). Therefore CI strategy may require enhanced ability to understand customer preference. Like Cole, Gershman (1991, chapter 3) emphasizes the importance of market presence with real products (versus prototypes) for learning how to design future products correctly.


This paper closes with a brief discussion of three research issues.

  1. Under what conditions is CI an appropriate strategy?
  2. As a knowledge-management strategy, how does CI differ from the usual iterative processes that organizations use to develop and test their products?
  3. How might design of experiments improve approaches for managing knowledge acquisition during CI?

First, what factors are critical for an effective CI strategy? Cole’s examples suggest that CI may replace discontinuous types of product improvement when cost of manufacturing and distribution are low, when economies of scale are not dramatic, and in rapidly changing environments when future directions of development are fuzzy. Scholars of innovation suggest research into the following: organizational factors like leadership characteristics (for example, attitude and support toward change); degree of centralization; complexity (for example, job diversity and professionalism); formalization and rigidity; interconnectedness (the informal social network that enhances diffusion); and organizational slack (the degree to which resources are available). STS and creativity researchers suggest additional factors.

Second, most Fortune 500 organizations that produce consumer products break the early part of their product development process into a network of small steps that test product concepts and prototypes, shifting and winnowing product and feature mixes to continually adjust for changes in market preferences and segment focus. Many companies have rapid-strike forces that address emerging strategic issues and rapidly initiate innovative products as new technologies, markets, and competition emerge (Liedtke 1997). How do these learning practices integrate into or differ from CI?

Third, Cole points out the link between CI and experimentation. Within design of experiments (DOE), the thinking behind CI has been part of quality for more than 80 years. DOE provides philosophy and methodology for efficient and effective knowledge acquisition by controlled introduction of variation and error (Box 2000; Box, Hunter, and Hunter 1978; Taguchi 1993, 1994). For example, conjoint analysis involves the design of combinations of products, product features, and prototypes to rapidly learn about consumer preferences when introducing innovative products and services to rapidly changing environments. None of the products in the mix designed for these market experiments may be optimal since the primary purpose of the design is knowledge acquisition for latter designs. Furthermore, the design deliberately introduces variation or error, as Cole calls it, into the design of these products for the purpose of learning. DOE also has technology for iterated and sequential design, where early designs (for example, screening designs) are probes that explore completely unknown areas and focus the variables of knowledge acquisition for learning in subsequent experiments. Sequenced DOE can suggest cessation points when the value of additional information diminishes.

This author believes there are many research opportunities to expand and abstract DOE philosophy and technologies to further develop CI. For example, EVOP is an evolutionary operational DOE strategy that automatically introduces variation (error) into the production process in order to learn how to improve products and efficiencies (Box and Draper 1998; Box 2000). EVOP suggests CI may also be a strategy for production innovation, not just product innovation.


Box, George E. P. 2000. Box on quality and discovery, edited by George C. Tiao, Soren Bisgaard, William J. Hill, Daniel Pena, and Stephen M. Stigler. New York: John Wiley and Sons.

Box, George E. P., William G. Hunter, and J. Stuart Hunter. 1978. Statistics for experimenters. New York: John Wiley and Sons.

Box, George E. P., and Norman R. Draper. 1998. Evolutionary operation. Reprint of 1969 edition. New York: John Wiley and Sons.

Deming, W. Edwards. 1986. Out of the crisis. Cambridge, Mass.: MIT Center for Advanced Engineering Studies.

GEMS. 1999. Six sigma in the new millennium—Learning session. Chapter 6. Lake Geneva, Wisc.: General Electric Medical Systems. America’s Quality Forum.

Gershman, Michael. 1991. Getting it right the second time. Boston: Addison-Wesley Longman.

Huber, Vandra L., and Karen A Brown. 1991. Human resource issues in cellular manufacturing: A sociotechnical analysis. Journal of Operations Management 10, no. 1:139–158.

Imai, Masaaki. 1986. Kaizen. New York: McGraw Hill.

Isaksen, Scott G., Roger L Firestien, Mary C. Murdock, Gerard J. Puccio, and Donald J. Treffinger. 1994. The assessment of creativity. Buffalo, N.Y.: Center for the Studies in Creativity.

Juran, Joseph M. 1992. Juran on quality by design. New York: Free Press.

Justin, Martin. 1995. Ignore your customer, Fortune, 1 May, 121–126.

Kano, Noriaki. 1987. TQC as total quality creation. Proceedings of the International Congress on Quality Control, Tokyo.

Kano, Noriaki, Nobuhiko Seraku, Fumio Takahashi, and Shin-ichi Tsuji. 1996. Attractive quality and must-be quality. In The best of quality, edited by John D. Hunt. Volume 7 of the Book Series of the International Academy for Quality. Milwaukee: ASQC Quality Press.

Liedtke, Charles A. 1997. Horizontal interactions during strategic improvement initiatives. Ph.D. diss., School of Business, University of Wisconsin–Madison.

Leonard-Barton, Dorothy A., and Walter Swap. 1999. When sparks fly: Igniting creativity in groups. Boston: Harvard University Press.

Pasmore, William A. 1988. Designing effective organizations: The sociotechnical system approach. Wiley Series on Organizational Assessment and Change. New York: John Wiley and Sons.

Rodgers, Everett M. 1995. Diffusion of innovations. New York: Free Press.

Taguchi, Genichi. 1993, 1994. Taguchi methods. Volumes 1–4. Quality Engineering Series. Dearborn, Mich.: American Supplier Institute.

Taylor, James C., and David F. Felten. 1993. Performance by design: Sociotechnical systems in North America. Englewood Cliffs, N.J.: Prentice Hall.


Mark P. Finster is an associate professor in the school of business at the University of Wisconsin–Madison and a contributing member in the Center for Quality and Productivity Improvement, the Consortium for Global Electronic Commerce, and the Center for Quick-Response Manufacturing. He also serves on the governing boards of the Center for Manufacturing and Technology Management and the Center for Manufacturing Systems Engineering. He holds the Gaumnitz Distinguished Faculty Award, the Mabel W. Chipman Excellence in Teaching Award, and is a five-time National Science Foundation Scholar. He chaired the establishment of NSF’s efforts to develop a national research agenda in areas of quality and has consulted with more than 100 organizations on four continents.

Finster earned a Ph.D. in mathematics from the University of Michigan. He may be contacted as follows: School of Business, University of Wisconsin, 975 University Avenue, Madison, WI 53706; 608-262-1998; E-mail:

If you liked this article, subscribe now.

Return to top

Featured advertisers

ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.