Coles Group Ai personalisation model launch
Tailoring a personalisation model for a Coles 6 million + weekly online shoppers
Making Australia's most recognisable brand more relevant, future-proof and exciting
The Client
Coles Group is one of Australia's leading supermarket, retail and consumer services chain. Coles online shopping platform uses multiple silos of data to gather real-time product information to be used in communication channels. Coles customers relies on 3 main emails weekly to track their shopping, promotions and updates. As Coles transitions to a new centralised system, so must all data pools needed to be combined into a single integrated platform that addresses two growing challenges: Relevancy of Products to customers and Website management.
The Challenge
Coles Supermarket devision wanted to appeal to enterprise prospects by emphasising the need for customer experience. Meaning Coles has to provide a new website experience that can handle the purchase requests, talk to new logistics data-points and to provide customers a grid of products pulled from promotional lists, ensuring they are available and in stock which is updated in real time. The weekly communications also needed a refresh to provide customers with products that are relevant for them. The website needed to also allow active customers to save products and browse products that are suggested for them.
The Solution
After many consultations, workshops and meeting with Coles leadership across divisions, we pulled together 25 use cases that allowed a new personalised aggregator model to suggest products to customers based on their in-store and online purchase history. We needed to develop a model that included identifying the right customer cohort which can be updated for the entire group and tagged in a way that we can building learnings from for future use cases. In the end we began rolling out a series of emails featuring products that were on promotion from a weekly catalogue that customers can shop and save for later instances. The result of over a year of work showed that the new model not only increased engagement by a staggering 35% higher open rates, 9% click through rates but an increase of $4million dollar uplift week on week for online sales over the test period.
