FAQs Frequently Asked Questions
Here are the most common questions about foodpairing®. If you didn’t quite find what you were looking for please use the search field.
What is an AI Agent?
An intelligent assistant that executes specific workstreams for your team.
What tasks can the Foodpairing Concept Agent do?
Turn marketing briefs into validated product ideas.
What is Headspace?
The platform name where you can enter instructions or tasks to the Concept Agent.
Can I feed the Concept Agent with specific prompts of my marketing brief with detailed parameters (brand, target audience, occasion, opportunity space)?
Yes. You can provide anything from a short 1-2 sentence description to a full marketing brief, trend reports or consumer insights. The more detailed information you provide, the more accurate the output becomes.
What exactly is the output of the Concept Agent?
Each product concept generated includes a product image with brand elements, product title, main ingredients, reasons-to-believe why this product is on-trend and consumer relevant, overall liking score, purchase intent prediction and a product concept description. Try our Concept Agent for free.
In which use-cases is a Concept Agent useful?
The main use-case is translating marketing briefs into product concepts. Other use-cases are rapid product ideation, brainstorm tool, fueling product pipelines, rapid prototyping, fast consumer insights and validation, creation of flavor variations, creating briefings for product developers, inspiration for product development, etc.
How are the product ideas validated?
A Digital Consumer tastes the product ideas virtually and provides a predicted Liking and Buying Intent score.
What is a Digital Twin or a Digital Consumer?
It is a replica of a target consumer. Similar to traditional ‘Target consumers’, a digital consumer has demographics like age, gender, country. Attitude data like openness for innovation or variety seeking behaviour. And finally category behaviour like frequency of purchase, preferences, likes and dislikes.
Example: 18-35 years old, male, from US, who purchases snacks twice a week, prefers low cal snacks, opened for new flavors, and likes brand X.
How do you build a digital twin?
A digital twin is an AI model trained on survey data from real consumers. The data of the survey is used to train an AI model to “taste” exactly like those real consumers. The digital twin IS that survey data transformed into a predictive model.
What does the survey look like?
Within the survey we ask real consumers to score (0-9) of about 100-150 ingredients/products to understand drivers of liking across all categories (e.g., savory snacks, drinks, confectionery, etc.).
What is the minimum sample size to create a digital twin?
500 real consumers.
Do consumers taste actual products when you create digital consumers?
No. Digital twins are created purely from survey data – no physical product tasting is involved.