Finding the right coffee maker for Walmart
Role
Product Designer
Skills
Figma, Miro, User Research, Prototyping
Project Length
6 Months
Background
Eko specializes in interactive video experiences. I was hired to work on their Walmart account to help create a platform that guides their customers to make better decisions about their purchases.
The first vertical they decided to explore was coffee makers, which would create the architecture for future verticals. The first brand they partnered with was Keurig.
Goals
Defining goals was important to the success of the project:
Walmart had three main goals:
Help people find the best product for them
Drive more traffic to their PDP pages
Reduce cart abandonment
Eko had one main goal:
Drive as much traffic as possible, the main metric for this project was conversion from our experience to their Product Detail Page.
Customer
The customer in this case was anyone who was shopping for a new coffee maker, and was unsure of which one to pick.
A small problem with big impact
Walmart provided the initial user research, and numbers. Cart abandonment was higher than industry average for both desktop & mobile.
Those small differences in percentages can mean millions of dollars as the coffee market in the US is ~$7.55 Billion annually.
| Abandonment on Desktop | Abandonment on Mobile | |
|---|---|---|
| Industry Average | 69.57% | 85.65% |
| Walmart Coffee Makers | 71.23% (1.66% above standard) | 88.1% (2.46% above standard) |
Goals for Improvement
Walmart had a multi-tiered approach to increase the numbers, and Eko was in the “filling the funnel” part. I was tasked to help increase the percentage of conversion from our experience to their PDP page.
| Conversion Rate | Desktop | Mobile |
|---|---|---|
| Current | 2.1% | 0.9% |
| Goals | 3.0% (Increase of 0.9%) | 3.0% (Increase of 2.1%) |
User Research
I conducted eight in-depth interviews using a prototype planning to buy a new machine in the next 3 months. Some criteria for the research participants:
30-45 year old home coffee drinkers
Planning to buy a machine in the next 3 months
Walmart Shoppers
Shopped online at least once in the past month
Key decision maker/purchaser for the home
Learnings - The Good #1:
Experience Seems Helpful at the Beginning:
✅ Narrows down options
✅ Saves me time
✅ Allows users to start the process from home
“It’s a good starting point. It can guide you through... [when] you don’t know what you’re looking for. ...It gives you some information right away, and a place to start.”
Learnings - The Good #2:
✅ Feels more ‘guided’ than other online tools:
✅ Provides some helpful tips and advice
✅ The host seems credible
“I thought it was easy and informative without being overbearing. She’s not forcing me to make a decision but helping me to make an informed decision.”
“I liked that it gave me a bit more information to guide me towards what type of coffee maker I should choose.”
Learnings - The Not Good:
❌ 5/8 Participants though it was too ad-like and not catered to the experience
❌ People generally wanted more choices
“At first I thought it was an advertisement. It made me feel like I hit a dead end.”
“I was confused that I only saw one suggestion at the end, I was expecting to see a range of choices. ”
Iterations
Finding the correct amount of recommendations was important
I ran a series of smaller tests of 3 different versions determine the correct amount of products to display.
Iterations continued
The animations were important for the premium feel.
They also helped seamlessly transition the user to the shopping experience.
Final Experience
Impact of Updated UX
After a few more versions, the effort was worth it, and the desktop version went above our goals.
We didn’t quite hit the results on mobile, which was the next point of focus.
| Conversion | Desktop | Mobile |
|---|---|---|
| Final Version | 3.5% (0.5% above goal) | 2.6% (0.4% below goal) |
| Goals | 3.0% | 3.0% |
Lessons
It is good to try out quick & low cost experiments, but it is important to know when it is time to connect with users.
It is very beneficial to have user testing candidates who meet all of the criteria of potential users.
Next up