Morrisons reaches 50,000 daily searches with Gemini AI in-app product finder built on Google Cloud
UK supermarket chain Morrisons launched an in-app Product Finder powered by Gemini AI models on Google Cloud — using BigQuery and Cloud Run — to help shoppers instantly locate items by aisle and shelf position. The tool saw 50,000 daily uses during the Easter 2025 peak period, reducing shopper frustration during seasonal store rearrangements.
Background
Morrisons serves 10 million shoppers weekly across the UK. A recurring frustration for customers was the inability to quickly locate specific products, particularly during peak seasons when stores undergo temporary layout changes for promotional displays. The supermarket needed a low-friction solution that worked within its existing mobile app and could draw on real-time inventory location data across hundreds of stores.
What Was Implemented
- In-app Product Finder built using Gemini AI models on Google Cloud
- Cloud Run functions process customer natural language product queries
- Gemini models interpret queries and correlate them with Morrisons' internal product code database
- BigQuery houses comprehensive product data including availability and precise aisle/shelf location for every Morrisons supermarket
- Results returned to the customer in seconds, providing real-time location accuracy
- Deployed within the existing Morrisons mobile app — no new hardware or app required
Results
The Product Finder reached 50,000 daily uses during the Easter 2025 period, a peak time when temporary seasonal layouts increased the difficulty of finding products. The solution provides real-time accuracy for product location across Morrisons' network of 500+ supermarkets. No pre-deployment baseline metric or conversion/satisfaction outcome was disclosed in the announcement; the 50,000 figure is a usage volume metric only.
Lessons
- Centralizing product and location data in a cloud data warehouse (BigQuery) enables rapid AI-powered application development at retail scale
- Solving a single, high-frequency customer friction point ("where is this item?") can drive immediate and measurable engagement
- AI-powered product location is especially valuable during seasonal store rearrangements when staff are unavailable to assist every shopper
- A joint vendor-retailer press release provides usage volume but typically omits baseline comparisons; independent outcome metrics (e.g., customer satisfaction scores, conversion lifts) should be sought in follow-on reporting