prediction markets Archives | The Garage Group

prediction markets

When it comes to product concept screening, most traditional research methods are pretty cut-and-dry: Get some standard metrics (purchase intent, purchase frequency, etc) from individual consumers in order to build a forecast model to predict product success in the marketplace. However, this method is usually expensive, time consuming and not conducive to iterative idea development. And, they’re not very engaging for research respondents.

Enter prediction markets.

Prediction markets take the answers of the crowd into account. In his book, The Wisdom of Crowds (2004), James Zurwecki posits that, “under the right circumstances, groups are remarkably intelligent, and are often smarter than the smartest people in them.” (This theory rings true in a lot of familiar contexts today, from search engine design to the ups and downs of Wall Street.) Prediction markets allow those “right circumstances” to happen by picking a tailored group based on exactly what the researcher is looking for—and by allowing individuals to play off the collective voice by seeing how their opinions stack up against the rest of the group (as opposed to taking a static, mono-faceted survey in a vacuum).

Prediction markets help marketers and product developers understand the relative strength of a concept by effectively bridging the gap between corporate research methods (predominately based in quantitative data and a large sample size) and startup research methods (predominately based in iterative exposure and smart projection). We like prediction markets since they are so consistent with our philosophy of gathering quick, high-quality insights for truly breakthrough results. And, they’re as good, if not better than traditional methods at identifying top ideas.

Here’s how prediction markets work: Consumers are asked to “predict the success” of various concepts or ideas (or really just about any stimulus). They predict just those items about which they have some knowledge (so, someone who doesn’t own a dog mostly likely wouldn’t predict the success of a new dog food idea). As respondents choose ideas they think are more likely to be successful, they assign a number of tokens — allocating tokens based on the ideas they believe are the strongest. As they see additional ideas, they can re-allocate their tokens. It’s a fun, engaging way for consumers to sort through ideas, pick the strongest (relative to each other) and then weigh in on the strength of individual ideas using their tokens. They see how the rest of the “market” (i.e. other respondents) are investing their tokens, and again, can re-allocate. It’s like real-time investing, and the learning is invaluable. Open-ended questions capture qualitative insight into why respondents placed their tokens on specific ideas.

It’s an engaging, gamified platform that encourages respondents to stay involved and focused on the task at hand until it’s complete, yielding more thorough, thoughtful insights.

1. Prediction Markets are Flexible.
Test ideas at a much earlier phase, using more robust input than traditional purchase-intent based concept test methods–and iteratively develop and then re-test ideas even as the idea format changes.

2. Prediction Markets are Fast.
Done within days (as opposed to several weeks of development and field time for traditional methods).

3. Prediction Markets are Affordable.
Inexpensive when compared to other concept-testing approaches, allowing teams to plan for iterative rounds of testing as they learn and optimize ideas.

4. Prediction Markets are Accurate & Insightful.
They not only tally responses, but also take into account the confidence the respondent has in his/her answers. Teams are able to identify the big winners as well as weaker ideas early in the process.

For big companies looking to innovate more like startups, prediction markets are a great first step. After all, they closely model the proof-of-concept process that entrepreneurs use in real life. Entrepreneurs typically present an idea to potential investors/stakeholders, who then must decide whether or not to put money on the concept (just like betting on ideas with tokens in a prediction market). Startups also ask for feedback from multiple perspectives before taking the plunge to execute on an idea (just as prediction market surveys do). The Shark Tank-style competition in The Garage Group’s ideation process–where “consumers” get $100 to invest in their favorite ideas–is actually a great example of the prediction market process come to life. It’s an extremely effective activity within our workshops to help clients determine standout concepts. Prediction market testing engages consumers in the same process.

Have you used prediction markets as a part of product ideation? What did you think of the process? Did it get you the results you needed?