Recommendation Engine

We worked with Abbott Pharmaceuticals to complete a product affinity analysis – identifying which Abbott products are most likely to be purchased together. We used this intelligence to inform Abbott’s sales, account based marketing plans and online user experience.

Our approach used a collaborative filtering technique. This enabled us to identify affinity between products as well as next best product recommendations. And to ensure that our models worked, we validated them against historic transactional data – a step that showed 81% accuracy in predicting next product purchases.

The model is integrated into Abbott’s operational processes and its accuracy is increasing as it ‘learns’ from more data and further updates.

Client

Specialisms involved

We worked with Abbott Pharmaceuticals to complete a product affinity analysis – identifying which Abbott products are most likely to be purchased together. We used this intelligence to inform Abbott’s sales, account based marketing plans and online user experience.

Our approach used a collaborative filtering technique. This enabled us to identify affinity between products as well as next best product recommendations. And to ensure that our models worked, we validated them against historic transactional data – a step that showed 81% accuracy in predicting next product purchases.

The model is integrated into Abbott’s operational processes and its accuracy is increasing as it ‘learns’ from more data and further updates.

81%

81%

Accuracy in predicting next product purchases

IRIS NEWSLETTER

Be the first to hear what we’re up to

We use cookies to make this website work better for you and to track site visits anonymously. You can opt out if you like.Tell me more