Improving Ocado’s

Checkout Walk

TL:DR

The Checkout Walk on Ocado.com was a series of 6 pages shown after the user presses checkout. It was commercially very successful but a poor user experience and needed refreshing. Through a diary study, previous research and page analytics, I was able gather insights and propose a new experience that was a single page long, utilising carousels for various categories of product, with personalised recommendations at the top of the page.

How did we approach it?

A 3-pronged methodology was used:

  • A diary study of existing users, who recorded their whole shopping experience - to create a realistic attitude towards the checkout flow - and then spoke out loud during the Checkout walk pages.

  • Commercial analytics to determine average additional spend on each current page.

  • Previous A/B testing research that swapped the page order of the Checkout out flow to determine if positioning or offering had the most impact on profitability of the page.

What were we trying to understand?

The Checkout Walk on Ocado.com was a series of 6 pages shown after the user presses checkout. It was commercially very successful but a poor user experience. I aimed to understand what aspects of the walk were liked by users and where friction was felt in order to guide our designers to build a new experience that was better for the user but did not damage the commercial impact of the page.

What did we learn?

  • The checkout flow does annoy users particularly on pages where it offers non-personalised products e.g. offers on meat products presented to a vegan user.
    More personalisation -> Less user frustration.

  • The flow is too many screens and requires a lot of scrolling on mobile devices to move forward.
    Limiting the number of screens is key.

  • The first page positioned performs best commercially.
    A single page view should retain the commercial aspects the business requires.

  • “Have you run out of” a page that displays products the customer has previously bought performs best commercially (regardless of positioning).
    Personalisation is key to user engagement both attitudinally and commercially

  • “Flash sales” was a popular page of pot-luck items, even if it wasn’t personalised.
    Keep in flash sales.

  • The ability to use one-click on the mobile app meant you could skip the Checkout walk entirely.
    Users liked when they found ways around the flow - we need to streamline the flow.

Impact

The data gathered was able to give us a clear steer on what content we could remove from the Checkout Flow (non-personalised goods) and what aspects should remain (personalised suggestions and flash sales).

The mixed methods approach utilising user data, quantitative analytics and commercial information meant the stakeholders were able to understand the inter-play between the user’s preferences and the commercial performance of the flow, allowing for a solution we were confident served both functions successfully.