Editor’s Intro: Concept testing has been a core practice within marketing research for generations. Yet, despite all that experience, most new products fail. Over the past few years, we have seen new methods introduced that focus on behavioral reactions to new ideas, rather than attitudinal reactions. A/B market testing and prediction markets have been used with success. Ranjana Gupta describes another method that several firms have started working with – crowdfunding.
Circa 2000. An undergrad has the idea to create an online social network. To estimate its potential, he asks 100 people to rate its likely usage on a 5-point scale.
It is likely that the top box score would not have been high.
Researchers would protest that he needed matched panels. But how do you benchmark the unbenchmarkable? Market research could have killed Facebook.
The answer is crowdfunding:
The undergrad has an idea but no money. What if he pitches his idea to everyone on campus and asks them to invest? He gets the ‘crowd’ to ‘fund’ him.
That is the crowdfunding premise. It deals with limitations of old MR:
- ‘Innovation’ benefits are not clear to an average consumer – get rejected in direct questioning.
- Respondents do not commit to intention – average ideas do well.
- Consumer choice is influenced by what others do – this cannot be assessed in panels.
- Crowdfunding balances System 1 and 2, allowing people to react naturally in an organic environment. The results are observed behaviour based and not claimed intention.
Works on 2 principles:
- Investment is intention practiced: People pledge money for ideas they are interested in using themselves.
- People influence people: we are influenced by what others do and say and also compare one idea with another.
These principles were tested in collaboration with a market leader in India. Concepts from various categories – tech, F&B, grooming – were uploaded on an actual crowdfunding site in a true competitive context.
People were invited to spot the next big innovation and pledge money towards the development of any ideas they liked. In return, they would be the first to receive the end product. End to end behaviour of each visitor was recorded. (Here is a demo of the site.)
The central deliverable is the Prospect Matrix formed of 2 metrics:
- Engagement: will the concept generate interest? – the number of concept visitors as % of total visitors.
- Action: composite score derived from money invested and the number of investors.
For both metrics, the maximum value is 100; higher is better.
Meaning of each quadrant:
- Top-right (Engagement: High, Action: High) is for Champions. These are concepts/ideas that are likely to be crowd pleasers.
- Bottom-right (Engagement: High, Action: Low) are Middling, me-too products. The concepts here garner casual levels of interest but do not have anything too special that call people to action. Me-too products are generally seen here.
- Top-left (Engagement: Low, Action: High) is for Slow-starters or niche ideas. Concepts here are likely to appeal only to a certain cross-section of the intended audience, at least initially. They are potent enough to gain popularity and a critical mass with time.
- Bottom-left (Engagement: Low, Action: Low) is for Laggards. Concepts that are not likely to fare well are present here.
The matrix indicates concept trajectory, replacing the top box led pass/fail system of conventional tests.
Traditional testing for the same concepts showed similar potential for common ideas, with nearly 100% positive scores while the more disruptive ideas like the AI enabled personal grooming product was rejected. In contrast, Engagement and Action scores from crowdfunding were low for the best performing concepts in a traditional test. Most telling was the position of AI enabled personal grooming product in the Slow Starter quadrant of the Prospect Matrix, indicating potential and not rejection as seen in the traditional method.
Isn’t this similar to prediction markets?
It taps into the power of crowds like Prediction Markets. Prediction markets are futures markets “run for the primary purpose of using the information content in market values to make predictions about specific future events”. True prediction markets rely on the expertise of the individual making predictions and are used extensively in economic and political areas. In MR, the principles are adapted and simplified to allow regular consumers to predict what would be the outcome of certain marketing choices, like “XYZ is planning to launch a new shampoo that needs to be used just once a month. Do you think this product will be a success in the market?” The consumers bet on the outcome (usually binary in nature, like ‘Yes / No’, using certain chips or equivalent allotted to them. They can also see what others are betting on. The value of their own bets increase depending on whether they are on the winning side or not.
Use of crowdfunding differs in the following ways:
- There is no competitive context in most Prediction Markets
- There is no betting in crowdfunding; in the adapted form being discussed here, people pledge money with the motivation of getting the product/service they like to the market as soon as possible.
Since the first pilot, we have tested the approach in markets like US and UK and in different categories; the results have been remarkably consistent, nuanced and logical. It validates our belief that there is power in crowds; we just need to harness it to innovate ourselves and for our client partners.
 Berg and Rietz, 2003