Dunelm

Situation:

My role as Senior Product Designer was to help identify customer friction points and run A/B and Multivariant experiments in order to increase checkout conversion.

I had successfully launched a re-design of some core customer pages from within another team and moved into the Cart team to optimise their Cart to Checkout conversion rate.

As the Lead Product Designer on the Cart team, I began looking at current analytics data to determine what percent of customers convert from Cart through to Checkout. During my analysis, I discovered that Cart to Checkout conversion was extremely low compared to the Cart page traffic.

To start, I analysed ContentSquare and Google Analytics funnels to determine what percent of customers convert from Cart through to the Checkout flow.

Hypotheses:

I created 3 key hypotheses as to why we were seeing such a large drop in conversion through to Checkout and presented them back to my team with a solution proposal.

1. Fulfilment options are disengaged from the Checkout button which confuses the user. My user interview research showed that users were expecting the fulfilment options to progress to checkout.
2. During user interviews I found that Error messaging is too conversational and not instructive. Users found the messaging too confusing them as to which items the message referred to in the cart.
3. When a user has items in their cart that are or need different fulfilment methods, they get confused since the fulfilment options apply to the entire cart.

Solutions to Hypotheses:

I validated each hypothetical solution with User Testing (UT.com), Surveys (Lyssna) and Exit Surveys on Cart pages as well as analyzing analytic funnels I created to isolate these multivariant tests.

1. I added fulfilment options above Checkout button
2. Added messaging at item level
3. Grouped fulfillment options at item level

Result:

I released each of the 3 variants into production for 50% of our customers broken into 3 challengers at 33% for each variant. The plan was to measure the success over a month. I monitored Screen Recordings as well as analyzed success metrics such as:

• conversion rate to checkout
• time on page
• Average Order Value
• Revenue per Visitor
• Bounce rate
• Pages per Session.

I also managed to integrate multi session cookies so that I could track user behavior over return visits to the Cart. I wanted to know where they entered the site before they landed back on Cart page and where they went afterwards.

After determining that conversion from Cart to Checkout had not only stabilized but increased 3% to match the new increase in Cart traffic, I released the winning variant into production at 100% along with the re-design and then repeated the process to introduce each of the other 2 variants slowly into production based on their successful outcomes. This was to ensure that the variants worked together on the Cart page and not just in isolation.