CONTACT US FLOW

Wayfair

2019

ABOUT THE PROJECT

Purchasing furniture online can be challenging. As consumers, we like to touch fabrics, sit on couches, lay on mattresses, compare colors in real lighting, the list is endless. As much as we want to give our customers the best experience while shopping for their home, incidents happen. Items can get damaged during delivery, packages can get lost, the hardware pack for assembly may be missing, or you just don’t like how it looks in your room. To date, Wayfair regularly sees more than 60K calls and 15K emails daily due to post-purchase related incidents. This project is all about giving our customers a quick and smooth experience to solve their issues and, as an outcome, reduce costs associated with these contacts.

PROBLEM

The existing flow had an abandonment rate of 26% at the time I joined the project. We started by understanding the root causes and found through user and quantitative research that many customers were not looking for issue resolution, but only seeking information. This information could be found in our Help Center page, but customers were contacting to get these answers rather than searching for them. Research also showed that customers hate calling, and those in need of a resolution believe that they need to call to solve their problems. 

Existing 4-step contact flow:

PHASES OF THE PROJECT

DISCOVER

We kicked off the project by defining goals and success metrics. Our overall goals as a team was to proactively answer customers’ questions before they contact customer service, create a SPA (single page application) for faster loading times in order to increase completion rate, and promote self-service tools to convert contacts into automated resolutions. The success metrics we would be measuring were the increase of flow completion rate, the reduction of the abandon to contact rate and the increase of self-service adoption.

KPI TRACKING

I began by creating a spreadsheet to understand and keep track of the existing flow’s KPIs. This helped me to keep track of the variations month to month and what design and content changes had been made each month.

This was tons of fun because it got me familiarized with the internal Wayfair analytics tools.

QUALITATIVE RESEARCH

I then proceeded with compiling all user research that had been previously conducted around contacting customer service and discovered 2 main trends: 1) There are many preconceived notions based on past experiences where customers are reluctant to solve issues on their own and felt like calling was the only way to get answers; 2) There were tons of confusions with the existing experience where we were not being clear on the process, causing more unwanted contacts.

JOB STORIES & HYPOTHESIS DOC

To keep track of all these findings and hypotheses, I created a spreadsheet outlining the problems we were trying to solve with job stories and our hypotheses. This document was shared with the participating team and the product manager used this document for roadmap planning for the coming quarters.

FLOW DIAGRAM

Next, I created a flow diagram to understand customer behavior and came up with hypotheses and questions for analytics to further investigate. Side note: it was interesting to find that there was a lot of missing click tracking data (e.g. where customers were landing from and null recordings).

DEFINE

After an extensive discovery phase, I realized this project was going to need a content first approach. I asked for a content strategist partner to co-lead this phase with me and together we planned a week long design sprint to create a new flow.

This is what our design sprint looked like...

✓ Kick-off with PM and tech lead to align goals

✓ Definition of 20+ customer use cases

✓ White boarding session led by my awesome content strategist partner Phaea Crede to define conversation-first approach

✓ Check-ins/feedback sessions with PM and tech lead

✓ Wireframing (me) and content matrix creation (content strategist)

✓ Final designs and prototyping - ready for user testing

DESIGN & DEVELOP

I simplified the selectable cards in order to have more flexibility for adding options to the flow and not have to depend on the branding team to provide illustrations for each option we wanted to add. This also made the mobile version more scaneable.

We added a section below each action with answers to the most common questions we receive through phone calls in order to reduce contacts for information only purposes, as well as convince the most contact oriented customer that they can do it themselves online and it’s not necessary to call. These messages are followed by the self service option as the primary CTA on the page with the secondary option “I still want to talk to someone” that opens the available contact methods.

RESULTS

The new design was tested on November 2019 with immediate positive results:

• Completion rate increased by 16.5%
• Abandon to contact rate decreased by 8.1%
• Self-service resolution adoption increased by 4.7%
• Also, the new design drove an unexpected 1% increase in phone sales

After testing was complete, the new design was launched to 100% Wayfair customers in December of 2019. In 2020, the entire Wayfair site went through a redesign we called 2020 Vision where this flow was also updated to the new style guide implemented across the site.

PERSONAL LEARNINGS

This project is very special to me because it was the first project I led when I joined Wayfair. It allowed me to familiarize myself with Wayfair’s internal analytics and customer service tools.

I was able to highly influence the end results from all the insights gathered from the discovery process (especially with the numerous calls I listened to) and built a strong collaborative relationship with the product manager, content strategist, engineers and data analyst.

I had a few ups and downs throughout the project, one down not being able to AB test a different order of the flow, an idea I had backed from data and user research. Because of some tech limitations, we had to stick to the current 3 option entry on step 1. We had to give in and work with that limitation and ended up having good results in spite of not being able to modify the order of the flow. Most importantly, it was fun! I made friends along the way and am proud of the journey we went through to have what we have in production today.