Finally, robust organizations take great care in building transparent evaluation and screening protocols. Openness reduces confusion about how ideas are evaluated, resulting in a participatory process. The protocols are made known in advance to all employees, Moreover, employees are solicited to serve as reviewers of ideas, advocates for ideas, and even to help peers in the crafting of ideas. This is a stark contrast to brittle organizations, who often promote secretive and selective evaluation protocols. In these organizations, ideas are promoted and advocated for based on the egos of the ideas creators, the preferential treatments provided by managers, and other nuances. Brittle organizations thus frequently create black-box innovation processes that discourage the submission of ideas and reduce belief in the merits of screening and advocacy processes.

Experimentation

Once an idea has passed through the screening process, experimentation and prototype-building begins. Even if the idea generator(s) or an advocate has created a prototype as part of an earlier stage of innovation, the idea's applicability to a specific problem, context or production chain must be tested. Experimentation does not test an idea's objective merits, but the suitability for a particular organization at a particular time. Ideas that are generated, advocated for, screened, and even found desirable might be ahead of their time or beyond the present capacity of the company. At either the screening stage or the experimentation stage, ideas may be set aside into an idea bank or idea library for development at a later time.

For instance, when the notion of redesigning the look and feel of bank interiors was raised at Washington Mutual, it was deemed worthy of experimentation. Washington Mutual then designed and implemented a new design in a few branches. That new style, named Occasio, was based on the open, friendly feel of many retail stores, and moved bank tellers from behind tall counters and into the same space as customers. Only once response to the new design was favorable did Washington Mutual then decide to revamp older structures and to apply the Occasio design in new branches.

Experimentation is an iterative process of development, and at each stage an idea may be re-evaluated by advocates and screeners alike. Because of this, experimentation may be continuous or may occur in fits and starts, depending on the presence or absence of advocates, screeners, and resources. Tested ideas are evaluated for suitability within a particular context and either are set aside or enter the diffusion and implementation stages. Experimentation results in data, prototypes, or feasibility studies that top management can use to evaluate an idea or innovation. Possible ideas are identified. The major outputs of this stage are libraries of ideas for the future and the identification of suitable ideas for commercialization. Another output is work-in-progress prototypes. Evaluating work-in-progress artifacts can point to new ideas that might trigger newer cycles through the innovation process.

Critically, resources must be present to experiment, and those in charge of the process must be given time to do so and time to reflect on the experiments so that refinements and evaluation are ongoing. This need not be an internal process, and in cases of product development and consumer response testing, it may be outsourced (Kambil, Eselius, & Monteiro, 2000). For business model, strategy, or service changes, however, this process must be internal. Experimentation can be resource-consuming, especially if an idea is found unsuitable or too expensive. Further, experimentation is an unstructured process that is difficult to make routine or familiar, since context is key.

Success indicators

Robust organizations tend to encourage experimentation, while brittle organizations emphasize the risk of failure and lack reward systems and structures that would encourage experimentation. The variety of ways in which robust organizations make it possible for employees to experiment with ideas in itself encourages experimentation. For instance, employees may be encouraged to write articles and attend conferences and trade shows to present ideas. Employees may also be encouraged to work with academic partners in formal arrangements to build and test ideas.

In general, robust organizations seek external constituents and their opinions, feedback and input for experimentation purposes. They thrive on taking in and understanding a multiplicity of interpretations. Robust organizations will also forge external partnerships to undertake risky and complex experimentations. For example, an organization might collaborate with a business partner or even a customer (e.g. a lead user) or even a research lab at an academic institution to experiment with a new product enhancement or service provision. Brittle organizations, on the other hand, do not know how to experiment with external entities. Much of this can be attributed to the fact that they have little experience sharing ideas with external entities or working on early-stage product development activities with outsiders. As such they perceive these activities as high-risk and dangerous, leading to an insular attitude on experimentation. The downside being that they very often have to assume all the losses with experimentation, leading to a negative connotation to the concept of experimentation, which in turn, negatively affects the innovation process.

For robust organizations, the process of experimentation itself is valuable and rewarding; even if the experimentation does not result in any products, organizational culture becomes further accepting of innovation processes. Brittle organizations lack these protocols and discourage what they perceive as risky, potentially-failed efforts at experimentation. Robust organizations do not view experiments lacking in results as failures, but as a learning activity. Learning throughout the experimentation phase is encouraged and sought after. After each run or trial, employees document their findings and connect them to original expectations and also to previous runs of the experiment. The output (i.e. the finished product or service), while important, is not the only measure of the value of experimentation exercises. One of the features of experimentation programs in robust organizations is the sharing of data and other artifacts that result from experiments. Robust organizations are more willing to share data between innovators so that mistakes are not repeated and a collaborative atmosphere to experimentation is developed. Moreover, sharing of data allows for hidden or latent connections in the data to the discovered through exploration of large data sets that have been collected over time and by multiple innovators. Today, especially in industries such as pharmaceuticals, biotechnology, and even financial markets, the sharing of data and the construction of large-scale data sets for experimentation is critical. Robust organizations have the capabilities and processes to foster the process of large-scale data collection and integration from experimentation. Brittle organizations, on the other hand, seldom have an organizational-view of experimentation. Each innovator is left to his/her own devices to experiment and hence are limited in their foresight and ability to solve complex problems.

Robust organizations identify ways to creatively use technology in the experimentation process. An example of this is the evolution of Goodyear's tire design and testing process. In the past, new tire designs at Goodyear were tested by subjecting the tires to road tests that included hours of driving around a track. This was slow and expensive and had an adverse impact on the environment. Due to the number of parameters involved, building a computational model to automate this testing process was perceived to be very difficult and challenging. In spite of this, Goodyear went ahead and developed computational models for testing tires virtually. The result was a faster, cheaper, and environmentally friendly process that used computer software to get the job done. Moreover, engineers at Goodyear were able to perform more tests and experiment more. Consequently, this led to the development of new products (Levinson, 2006).

Another example of a company where technology played a major role in enabling experimentation and thus helping innovation is JEA, a Florida-based public utility. JEA produces electricity using natural gas and oil. It implemented a neural-net-based solution to determine the optimal production of electricity by calculating the optimal amount of fuel required, based on the current oil and natural gas prices. The computer-based model also incorporated the government guidelines for prescribed emissions limits. The result was reduced operating costs with increased savings on oil and natural gas costs, and increased compliance with government emissions guidelines. Although the computer-based solution cost was significant ($800,000), the cost was recovered within eight weeks. Moreover, JEA is also planning to experiment with the system within its water business (Levinson, 2006).

The process, through which a prototype is constructed, tested, evaluated, refined, and improved upon, needs to be clear and known. Only a handful of organizations have sanctioned experimentation processes. Experimentation is most often conducted in a haphazard way by various individuals, teams, or organizational units, resulting in poor development of products and services. An apt analogy would be every scientist deciding to invent his or her own research processes and metrics. The growth of science would be stifled, as it would be difficult to evaluate, build upon, and even commercialize knowledge. The process of experimentation allows for the sharing of data, prototypes, ideas, and also past knowledge. Experimentation with ideas needs to be rigorous and in a manner that makes the ideas open to inspection (i.e. anyone who wants to know the process by which an idea has been developed should be able to examine the process). The experimentation process should be defined, yet not constraining. Not all experiments will be able to follow the outlined process exactly, and there might be developments in the environment that might warrant modifications to the process. Particularly in this stage, agility is crucial. Brittle organizations may have an experimentation process that is outdated, difficult to customize, and even cumbersome.

Print Get the Mag Weekly Updates