Using Vision AI and pre-trained AI agents allows to capture granular, in-the-moment product usage, context and experiences in self-made videos by participants to in-home product testing for example.
Here are the key learnings from the latest Ipsos POV on this approach:
WHY IS THIS SO INTERESTING?
- 70% of new product launches fail due to misalignment between a product’s promise and its real-world performance
- However, with the increased competition from smaller and local brands leveraging the impact of social media, global brands are constantly trying to compress product development cycles, thus reducing the time & resources dedicated to proper in-home product testing
HOW DO VISION AI AND TRAINED AI AGENTS TRANSFORM IN-HOME PRODUCT TESTING (or other observation-based research)?
- Self-made videos, by participants, capture the whole process of a product journey and experience
- With the right observational framework, videos reveal hesitations, small delight moments, unconscious workarounds, and environmental factors surveys commonly miss, and participants fail to remember afterwards when filling diaries or answering post-experience surveys
- An example of a rich framework Ipsos teams use: the one developed by anthropologist James Spradley, leveraging 9 dimensions to capture:
- For a better analysis of the video, developing AI agents with different specializations brings various perspectives enriching the human analysis and revealing subtle patterns a single human analyst might miss:
- Anthropologist
- Neuroscientist
- Product developer
- Contextual analyst
- Semiotician
- Behavioral Economist
To test this new approach, Ipsos teams have conducted a pilot study on the consumption of frozen pizzas at home. It revealed previously missed insights, like consumer frustrations related to the toppings that need to be manually rearranged before heating, or micro-expressions of delight and surprise at the product quality after first taste, that bring new ideas in packaging improvement and communication messaging.
CONDITIONS FOR SUCCESS?
- AI + HI: augmenting human intelligence, not replacing it
- Leveraging research and market experience with the right observational framework and the programming of specialized agents
Daniel Kahneman, the Nobel Laureate, differentiated between the “experiencing self” vs the “remembering self”: our evaluations of longer-term events (e.g., using a product) are primarily grounded in intense moments and the last experience in the event rather than the overall average (“Peak-End Rule”).
Leveraging AI in video analysis can help elevate the product and marketing development to a much richer consumer experience, bridging the gap between the experiencing and the remembering selves.