How we help the user to build up their first basket by using personalised recommendations
How we improved our first time customers experience
We kicked-off this project in September and we had a really tight deadline, as we wanted to help new customers during Christmas period, so we needed a quick turn-around before the peak time.
The problem
New customers to Sainsbury’s website typically have to build a basket of around 60+ items from scratch. What could we do to reduce this pain-point? We started to look at our data to discover more about our typical first-time customers. The highest percentage of new customer orders come from the ‘Executive Wealth’ Acorn group. These are wealthier families with good incomes who are frequent users of the internet. However, since the start of 2020, we’ve seen the greatest order growth from the ‘Student Life’ Acorn group, although they only made up 8% of total orders in August 2020 * (Acorn-user guide). Another group we wanted to help were our most vulnerable customers, who became critical to support during the first lockdown when we embarked on our Feed the Nation programme here.
Research and exploration
As a new customer you land on a website and it is like when you arrive for the first time to a new department store, you don’t know where the women department is, where homeware is or where the kids’ clothes are, etc. You feel confused and overwhelm so many signs and directions on how to find the product you want. Same happen when you are new to a website, we knew we needed to better support the new customers in several parts of the experience: the home page, search fields, sort and filtering, building their first favourites, to name but a few. With the time constraints ever pressing, we ran a scope meeting to narrow down what we could realistically deliver before December.
We thought it was important to understand the users’ pain-points when they come to the website for the first time and start building their basket. So we worked with this HMW statement: How might we help the user to build up their basket using personalised recommendations?.
In collaboration with a Google researcher and the Sainsbury’s research team, we ran a 250-user survey to understand how first-time users were building their baskets. From those findings and in parallel we created a few users’ stories that we tested with users. For example, we tested with customers several UI and different heading copy to identify which would be easier for the customer to use and make easier to understand for first-time customers. We also engaged with Google to improve the recommendations we were going to present users by incorporating an AI recommendation engine.
Proposal
We introduce new customers to their first favourite experience with products that are at popular to the majority of customer. The personalisation would happen over time as the customers use more and more the website and the AI will help to build that personalisation. We are learning from the current live experience by looking at the data and planning a couple of A/B test to validate a few other assumptions to improve the costumer experience and decide in the future of how we will roll-out the personalisation to other parts of the experience.
