Automating Product Return

Emma Sleep started in Germany and quickly grew into one of Europe's fastest-moving sleep-tech companies. They built their reputation on a simple USPs: a 100-night trial, a 10-year guarantee, and free delivery and returns. During the years I worked at Emma Sleep, the websites attracted an estimated 250,000-400,000 weekly visitors worldwide. In 2019, I created a self-service returns portal that saved customer service time, improved customer experience, and provided product teams with better feedback.

Problem

Emma Sleep offers a 100-nights-free trial for all their products for users, resulted in the soaring amount of return requests for customer service agents.

Emma Sleep's generous policies encouraged many customers to try the products with confidence. This also produced a high number of returns, especially during peak sales periods. The customer service team managed everything manually, which slowed down response times and left many return reasons undocumented. Product teams missed important insights because of incomplete data.

Goal

I set out to design a process that reduced workload, increased speed, and collected reliable data.

The aim was to relieve the customer service team, provide customers with faster resolution, and give product teams consistent return reasons for analysis.

Gathering Insights

I worked directly with the UK customer service team to understand the process.

I joined their daily work and processed return requests myself. I saw long conversations with customers about alternatives before a return was approved. I also observed clear spikes in requests during seasons like Black Friday and summer sales.

Gathering Insights

Less than 70% of returns included a documented reason.

The missing information limited the ability of product teams to improve the product. The most common reasons that were captured included "too soft" or "too firm" mattress.

Design Discovery

I mapped the full return journey and created a plan for automation.

I documented each step of the existing process and designed a self-service portal with stakeholders. The new flow captured reasons consistently and logged interactions automatically.

High-fidelity Design

Customers could log in with only their order number and email.

Fewer than 20% had created accounts at purchase, so the account-free design reduced friction. The flow also handled small items like pillows and bed covers more smoothly.

Solutions

The portal provided tailored alternatives for common return reasons.

For example, UK customers who reported a mattress as too firm were offered a free topper to adjust firmness. They also received an extended trial period to allow more time to adapt. Each major return reason was paired with a similar countermeasure.

Impact & Performance

The portal saved time for agents and reduced return volumes.

The easier process did not increase total return rates in the first three months. Customers value the options of trying a solution before sending products back.

-60%

In return rate for the reason "mattress is too firm" in the UK.Many users opted to try out a complimentary comfort layer instead of proceeding with the return process.

-18%

In return rate for other reasons. Due to the counter solution offered for various return reasons, we saw that many users have tried our solution before returning products immediately.

-20 hours

In weekly operating time for customer service agents.

Retrospective

The project delivered strong results despite resource constraints.

I worked within a two-person UX team that supported eight markets. Budget and time restrictions meant usability testing was done internally with employees instead of external users. For future iterations, I would run structured user interviews and external usability studies to validate the portal across a wider audience.

Gathering Insights

I worked directly with the UK customer service team to understand the process.

I joined their daily work and processed return requests myself. I saw long conversations with customers about alternatives before a return was approved. I also observed clear spikes in requests during seasons like Black Friday and summer sales.

Impact & Performance

The portal saved time for agents and reduced return volumes.

The easier process did not increase total return rates in the first three months. Customers value the options of trying a solution before sending products back.

-60%

In return rate for the reason "mattress is too firm" in the UK.Many users opted to try out a complimentary comfort layer instead of proceeding with the return process.

-18%

In return rate for other reasons.Due to the counter solution offered for various return reasons, we saw that many users have tried our solution before returning products immediately.

-20 hours

In weekly operating time for customer service agents.

Retrospective

The project delivered strong results despite resource constraints.

I worked within a two-person UX team that supported eight markets. Budget and time restrictions meant usability testing was done internally with employees instead of external users. For future iterations, I would run structured user interviews and external usability studies to validate the portal across a wider audience.

Gathering Insights

I worked directly with the UK customer service team to understand the process.

I joined their daily work and processed return requests myself. I saw long conversations with customers about alternatives before a return was approved. I also observed clear spikes in requests during seasons like Black Friday and summer sales.

Gathering Insights

Less than 70% of returns included a documented reason.

The missing information limited the ability of product teams to improve the product. The most common reasons that were captured included "too soft" or "too firm" mattress.

Solutions

The portal provided tailored alternatives for common return reasons.

For example, UK customers who reported a mattress as too firm were offered a free topper to adjust firmness. They also received an extended trial period to allow more time to adapt. Each major return reason was paired with a similar countermeasure.

Impact & Performance

The portal saved time for agents and reduced return volumes.

The easier process did not increase total return rates in the first three months. Customers value the options of trying a solution before sending products back.

-60%

In return rate for the reason "mattress is too firm" in the UK.Many users opted to try out a complimentary comfort layer instead of proceeding with the return process.

-18%

In return rate for other reasons.Due to the counter solution offered for various return reasons, we saw that many users have tried our solution before returning products immediately.

-20 hours

In weekly operating time for customer service agents.

Retrospective

The project delivered strong results despite resource constraints.

I worked within a two-person UX team that supported eight markets. Budget and time restrictions meant usability testing was done internally with employees instead of external users. For future iterations, I would run structured user interviews and external usability studies to validate the portal across a wider audience.

Design Discovery

I mapped the full return journey and created a plan for automation.

I documented each step of the existing process and designed a self-service portal with stakeholders. The new flow captured reasons consistently and logged interactions automatically.

High-fidelity Design

Customers could log in with only their order number and email.

Fewer than 20% had created accounts at purchase, so the account-free design reduced friction. The flow also handled small items like pillows and bed covers more smoothly.