FranceRetail and e-commerce2021 – ongoingMiam
Recipe suggestions that fill an online grocery basket
Miam's recipe tool matched ingredients to store products with 97% accuracy, and baskets with recipes ran more than 40% larger at Cora.fr
Problem
Shoppers often start a recipe and find they lack ingredients. Online grocers wanted a way to suggest recipes and turn them into full baskets. In early prototypes, only about 80% of ingredients were matched to products a retailer actually sold, which left items missing from baskets.
Approach
Miam, an eight-person French startup, built a white-label tool that retailers add to their websites. It suggests recipes based on shopping preferences and adds the ingredients to the basket with one click. Its own algorithms, run in a Python app, use Google Cloud's Natural Language API to read ingredient text and match it to products. The first retail client, Cora.fr, went live in February 2021.
Results
Takeaway
Matching free-text ingredients to real store products decided whether a recipe could become a usable basket. Miam reported the accuracy gain separately from the basket result.
Categories
- Size
- Small business
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- AI technology
- Recommendation and personalization
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Google Cloud's customer story describes how Miam built a white-label recipe tool for online grocers. It reports ingredient-to-product matching accuracy rising from 80% to 97% and an average basket more than 40% higher at Cora.fr when the basket contains recipes. The page carries no date; the Internet Archive first saved it on June 25, 2021.
Full details are in the original sources.
