Choicely: AI powered sustainable product finder

Insight

The majority of consumers say they want to shop more sustainably, yet time and again the data shows they don’t. When surveyed, many consumers say a barrier to changing their habits is understanding the options - sustainability is complex and confusing and alternatives are often niche or hard to find. We are creatures of habit and without a strong motivator, tend to stick with the easy and familiar.

Concept

To help overcome the green-gap and encourage customers to make more sustainable choices, we asked: what if we could help them discover more sustainable products in an easy and engaging way? It needed a simple and interactive mechanic that didn’t require customers to go out of their way to research or understand alternatives. And in order to prototype quickly, we needed a way to identify and match products without building a huge database of options.

Solution

To test the concept, we built a simple app that could take a customer’s product and use a pre-trained AI model to identify its category from just a photo or the product page. Using a combination of OpenAI's image analysis capabilities and Bubble's no-code platform, users can simply take a picture of any household product and receive a recommendation of a more sustainable brand. The whole prototype was completed from concept to launch in just 3 days.

The prototype app was built entirely with no-code tools including:

• Web scraping - Page2API
• Image analysis - OpenAI API pre-trained model
• Product category matching - OpenAI API file input
• Front end and integration layer - Bubble.io no-code editor