Increase conversions on premium clothing by addressing sizing concerns

Overview

Mariemur is a lingerie brand marketed only directly to consumers via an e-commerce store. The monthly traffic during our contract was 400,000+ users per month mostly paid Facebook and Instagram traffic.

Mariemur reached out to us, as they had reached a plateau in scaling paid traffic because a further increase in customer acquisition cost that was needed to increase traffic was no longer profitable. Also, the company couldn’t fully dedicate their in-house team to optimizing e-commerce store conversion and needed an experienced CRO agency.

Research

Based on funnel and quantitative user journey analysis, we discovered that the main funnel bottleneck was 84% cart abandonment. We began to investigate this issue, but event correlations, ABC analysis, and comparisons of buyer vs. abandoner segments didn’t give us representative data on the reasons for abandonment.

So we conducted an online survey of users who were about to abandon their shopping carts, with exit intent targeting. As a result, we found that 14% of users didn’t buy because they weren’t sure about the right size, and 82% of users in this segment didn’t know that Mariemur has a free return policy for the US market if the size doesn’t fit.

Combined segments of size issues were the second most common problem after price rejections (59%), which was not the area of our optimization efforts as discussed with the client.

Hypothesis

We proposed to overcome size rejections by adding “14-day money back guarantee. Free returns if size doesn’t match” for the US market and “14-day money back guarantee. Hassle-free returns if size doesn’t match” for other countries, since Mariemur only had a free return policy for the US market.
 

Validation of the hypothesis

To validate this hypothesis, we conducted an A/B test with personalized alternative versions based on user country. The experiment measured the impact of the new copy on conversion to transaction, cart abandonment, AOV and ultimately ARPU.

The experiment had the following characteristics:

Type of experiment: A/B test

Traffic split: 50/50

Device: Desktop

Primary metric: ARPU

Secondary metrics: AOV and CR

Version A – 2,296 users, 93 transactions, 4.05% conversion to transaction

Version B – 2,376 users, 124 transactions, 5.22% conversion to transaction