Checkout Optimization

OBJECTIVE: Improve conversion rate of checkout in the personal styling onboarding survey.

MOTIVATION: On the growth team, our goal is to continuously improve conversion rates.

MY ROLE: Senior Product Designer alongside a Product Manager, Product Marketer, Customer Service Leads, and 2 Engineers

MY ACTIVITIES: User Flows, UX / UI, Prototype Creation, User Testing, Story Grooming, Google Optimize A/B test

IMPACT: TBD

1 - How do we convince customers to complete checkout?

Dia&Co is a subscription box company that sends 5 items to a customer to try on at home and purchase. In order for them to receive those items, customers must complete a survey in which they tell us about their style, fit, and budget preferences. The stylists use that information to pick out the most relevant items for the customer.

Our Product Manager noticed that the each page of the survey had a high completion rate until checkout. This made sense because there is no commitment with a survey, but when it came time to order a box, some customers did not move forward. In order to increase checkout conversion, we needed to figure out why customers stopped moving forward in the first place.

2 - Use prototypes to understand customer behavior

The previous designer on the project created the first variant by deciding to separate the result page from checkout. She created a prototype of the existing experience and put it on a service called Validately, which screens and recruits candidates based on our criteria. It has them complete the prototype and discussion guide while recording the entire interaction for us to review later. Our team wanted to understand customer expectations and their reactions to the experience that we put in front of them.

The gist of the questions we asked remained the same for each of our prototypes, but changed slightly for each iteration as copy or page level content changed.

• What do you expect to happen when you click on the View Results button?

• What do you think the purpose of the View Results page is?

• What do you expect to happen when you click on the Get Styled button?

• Did you expect to see the Checkout screen?

• What remaining questions do you have about style boxes on checkout?

3 - Analyze prototype results

1st Round of User Testing - results

• Most believed that after clicking on "View Results" on the lead capture page, they'd see results about their body shape and see how we've styled women with that body shape previously.

• A number of users believed after clicking on "Get Styled" on the results page, they'd see sample clothing and outfit ideas on the next page.

• Most users were surprised they were taken to a checkout page after clicking on "Get Styled", as they thought they'd get more context on what the brand will "do for her" after clicking the button.

• By the time they reached checkout, participants wanted to know more about what comes in a style box and how the cost structure worked.

2nd Round of User Testing - results

• Participants believed they would see clothing options that match their answers from the quiz OR a model that fits her body type after click the “View Results” button.

• Majority of participants believed the purpose of the view results page was to see sample clothing and/or customized options for them.

• Participants expected to subscribe to a box of clothing when they clicked on the “Order a Style Box” button. They were not surprised to see the checkout page after clicking in the button.

  • “But if I just thought I was taking a quiz, I might be surprised that it was trying to sell me something. It would depend on how I got to the original page” - Participant

• Participant understood that the checkout page was trying to communicate that you can get a clothing box for a recurring $20 fee

• There will still questions when a customer got to checkout such as:

  • What does a sample box look like? What do I get in a box?

  • Do you pay later for what you keep?

  • How long are you allowed to have the Merch?

  • Is it possible to make requests or customizations?

3rd Round of User Testing - results

Majority of the participants answered the key questions in a way that demonstrated that they expected the experience that we had crafted, so I started designing flows for AB testing.

4 - Refine & Iterate Designs for A/B Testing

1st round of a/b testing parameters

• Baseline (current experience of combined results & checkout page) vs Variant 1 vs Variant 2

• 33% of traffic to each experience

Key success metrics

• Primary metric was conversion rate

• Secondary metrics were lead rate (filling out name and email) and checkout page views

results

• Baseline won

2nd round of a/b testing parameters

• Baseline (current experience of checkout) vs Variant 1 vs Variant 2

• 33% of traffic to each experience

Key success metrics

• Primary metric was conversion rate

2nd round of a/b testing - results

• Variant 2 won

5 - Future Iterations

3rd round of a/b testing parameters

• Baseline (old variant 2 - marketing copy) vs Variant 1 (utilitarian copy)

• 50% of traffic to each experience

Key success metrics

• Primary metric is conversion rate

2nd round of a/b testing - results

• TBD - utilitarian variant is ahead