Briefing

A recommendation engine seeks to learn from data and then filters, predicts and suggests items to a user that they may like to purchase or use. 

Most recommendation implementations are a hybrid of these systems:

1. Implicit: Uses machine learning algorithms to recommend products based on a user’s past interactions or those of users that have behaved similarly. Also referred to as the ‘collaborative filtering method’.
2. Explicit: Recommends products based on information that a user has provided, e.g. interactive content such as preferences, quizzes and product configurators. Also referred to as ‘content-based filtering’.

Meeting customer expectations

It’s been 10 years since Amazon shared that  35% of purchases were attributable to their product recommendations and Netflix estimated that 75% of programme choices were based on such algorithms.

With 71% of consumers now expecting companies to deliver personalised interactions and 76% getting frustrated when this doesn’t happen, adding personalisation to your website can increase sales, order value, user satisfaction and brand loyalty. 

Simply put, to meet this increase in consumer expectation, brands must utilise data and tech – like recommendation systems – to personalise customer experiences.

Getting started

To start personalising your customer experience with recommendations, follow these steps:

1. Integrate a tool on your website that offers an ‘off the shelf’ solution.
2. Study user behaviour and intent to understand the type of recommendation that would be most useful at different stages of the user journey.
3. A/B test your algorithmically powered recommendations to understand value.
4. Don’t limit your thinking to product recommendations, power other types of content recommendations to deliver a personalised experience.

Get in touch with our UX and CRO team to discuss your options and start personalising your website today.