How 25 people can be more effective than 400
When using the Swarm AI® platform to generate insights from human groups, we as market researchers are often asked about the small numbers of participants needed to get powerful results. “How can a swarm of only 25 people be more accurate than a survey of 400?” Here we review the core differences between swarming and polling, as they are extremely different instruments.
Active vs Passive
A critical difference with swarming is that it treats participants as active members of an interactive control system. Feedback loops enable the group to explore a set of options and converge on solutions that maximize their collective conviction. This active process is extremely different than polling, which treats each person as a passive data-point for statistical aggregation. Thus, while polling finds statistical averages within a group, swarming enables a group to explore a set of options. They can wrestle with the tradeoffs, and converge on the solutions they can best agree upon. This is why you may need 400 people, or even 4,000, when collecting passive data-points in a survey. In swarming only 25 people, randomly selected from the same population, are sufficient to converge on meaningful solutions. It’s a result of the control system, allowing them to reflect in real-time as they discover solutions that best represent the group.
Behaving vs Reporting
When conducting polls and surveys, researchers generally ask participants to self-report their sentiments. Unfortunately, studies show that individuals are highly unreliable when self-reporting their feelings 1, 2. Even if they are in-touch with their true sentiments, quantifying individual feelings in a form that can be aggregated across participants is very difficult, as every individual has different internal scales 3, 4, 5. This means the underlying data used by traditional polling methods are often highly distorted. Swarming, on the other hand, does not rely on participants to report their feelings; instead uses intelligence algorithms to track and processes how they behave while interacting as part of the real-time system. This behavioral data enables Swarm® to estimate the relative conviction that every individual feels with respect to the answer options.
Swarming is a unique and powerful method for harnessing the collective knowledge, wisdom, intuition, and insights of human groups, enabling diverse participants to quickly converge on optimized solutions with the support of real-time intelligence algorithms. While it’s tempting to compare the process to polls, surveys, and focus groups, the relationship is tenuous. Swarm® treats participants as active members of a real-time control system rather than reducing individuals to passive data points for statistical aggregation. The Swarm platform simultaneously evokes more authentic sentiments from individuals as they act, react, and interact with other participants. This enables the group to discover solutions that best represent their collective conviction rather than a crude statistical average.