In a provocative article published in Innovation Marketing in 2008, G Gimmy and M. Casabayo advise against consumer testing for innovation: “A team that justifies, to any degree, a request for funds with a consumer referendum should have its license to innovate withdrawn”[i].
This probably looks outrageous to most marketing and consumer insight champions; yet the reasoning makes sense.
They start with heart-breaking evidence: despite all efforts to professionalize innovation processes, FMCG products failure rate has increased over the years to more than 80%, while the proportion of truly disruptive innovation in FMCG has dramatically declined to just 7% (ProductScan, 2005).
They look at how consumer research is used in large FMCG companies in regards to innovation: consumer tests have a tendency to be used at best as a check-box to proceed to the next innovation process gate, at worst as a pure replacement to managerial decision-making. Gimmy & Casabayo compare the approach to “referendums” in modern democracies; votes of non-experts which override the rulers.
Finally, they look at the type of consumer research that is used: FMCG companies rely mostly on concept testing which fail to accurately predict success of truly disruptive innovation.
A paraphrase to Gimmy & & Casabayo conclusions could read: over the past 20 years ill-designed research has been used as a surrogate to managerial decisions. No wonder that managers are unhappy with their innovation ROI (BCG survey, 2006).
They leave us with some clues on how this could be improved:
- “Referendum” like testing should be avoided for truly break-through innovation
- Test protocols should be adapted to the specific innovation ideas
- Consumer knowledge should be put at the centre of testing rather than go/no-go criteria
- Innovation teams should be tested rather than ideas as they are the ones who can make things happen
While their advocacy against criteria-based consumer research might be a little extreme, the observations and the recommendations make a lot of sense.
On the consumer research side it is absolutely clear that traditional concept tests are a hurdle to game-changing innovation. First, because they do not capture the real potential of truly disruptive innovation; these ideas would typically be too “out-of-the-box” for consumers to recognize the interest. Secondly, because they do not capture environmental and social aspects of new products diffusion. Finally, because they are not effective at building knowledge over time: there can only be little ROI (in terms of time and money) in doing hundreds of one-off, low insight concept tests.
This idea that concept testing is not adapted to disruptive innovation is actually widely shared in the academic world. In her PhD thesis of 2006 on consumer research in the early stages of new product development, E van Kleef writes the following:
[…] The majority of available methods focus on evaluation of products. In these methods, products (ideas) are presented to a sample of consumers and evaluations are collected. These evaluations are used to optimise the product or to screen and select from different product ideas, ultimately ending up with the product idea with the highest likelihood of market success. However, these methods can be considered as reactive of nature in their use in the early stages. They constrain the researcher in the elicitation of unfulfilled consumer needs, because consumer input is restricted to responses to an already existing concept or product. A risk of relying on them solely is that they are likely to give product developers only ‘me-too’-ideas, which hardly excite the consumer. […] Most consumer research only attempts to build on existing and often already fulfilled needs of consumers. Consequently, the results of this kind of consumer research do not exceed common-sense knowledge and hence is consistent with what practitioners already take to be true.
The need to identify promising ideas and screen among alternatives is here; the real problem so far has been the lack of alternatives to concept testing/screening.
The challenge is how to fuel managers with foundational consumer understanding so they can make better decisions. And it is not necessarily the category specific understanding (which is typically well researched) that is needed, but the understanding of the persons; these human beings that will or will not adopt an innovation.
Yet, human beings are complex animals, so it’s not easy to know what to search for, to uncover these critical aspects of an innovation that may or may not make it a success. Cognitive psychology and behavioral economics have taught us a lot in regards to how decisions are made. This might be a good starting point to improve innovation research.
E. van Kleef highlights the following:
Understanding consumer behavior encompasses much more than just getting insight into how consumers evaluate and purchase products and services. […]Consumer behavior [can be defined] as all mental and physical activities undertaken by consumers that result in decisions and actions to pay for, buy, and use products and services. For consumers to decide to buy a product they must be convinced that the product will satisfy some benefit, goal, or value that is important to them. To develop a superior new product, consumer research needs to identify consumers’ product attribute perceptions, including the personal benefits and values that provide the underlying basis for interpreting and choosing products.
For example, it would be possible to research early adopters/opinion leaders (lead users research) and then research the product features that enhance diffusion. In 2003, wasn’t it those white earphones that made the success of the iPod? At first on posters, but soon after on adopters, who were highly recognizable with the product; a living marketing message.
For example, research could focus on the point of sales choice context, as this will be the first moment of truth for the new product. Understanding how people behave in that particular environment to trigger interest is probably as important as the technology itself.
These are only examples, and much more can be done in innovation research to escape from the “referendum syndrome”. The opportunity is huge, and the business results are likely to be as well…