DURATION

February 2020 - March 2020

MY ROLE

UX Researcher

DESIGN MEDIUM

Web Design

The Goal

Marmoset Music sells music licenses and produces original scores for creative projects. They wanted us to measure the success their recently launched site and how their online shopping experience + conversion rates can be improved.

I created an audit outlining the Marmoset Music online shopping experience. It consisted of UX research and test ideas + design recommendations for changes to make on their site to improve conversion rates. The live site can be viewed here. 

Heatmap Analysis

Hotjar heatmaps visualize interactions across the site from about 2000 sessions in aggregate. Using this tool I saw what users were interacting with on key steps in their shopping journey and what they weren't.  

Session Recordings

Session recordings are a great way to see bugs, issues, and obstacles faced by users on websites. I watched more than 300 sessions on Hotjar of the Marmoset site and picked out 2 painpoints that significantly impacted the purchasing process of many users.

Filter placement and logic may encourage users to overfilter on desktop. Selecting multiple subfilters within each category can immensely narrow down the criteria matching their search terms, and result in “Sorry, there are no results found matching your search terms”.

Resultantly, many users over filter while going through products, and are forced to unfilter and re-curate their products after this occurs.

Copy in licensing drawer is not scannable and density makes it hard for users to digest.

The current descriptions surrounding license use case scenarios, applications, and restrictions are very long and detailed.

The user in this particular recording uses their mouse to trace the lines of text as they read through it.

Unmoderated User Testing

Using UserTesting.com, I conducted usability tests on both the mobile and desktop experiences. I created different scenarios to mimic landing patterns of direct and organic traffic on sites, based on traffic acquisition data on Google Analytics. 

Homepage is engaging and speaks to brand differentiators.  Participants commented that they believed that there is a diverse music selection available at Marmoset, and that the moods + mixtapes spoke to the variety they have to offer.

Unique music filters create excitement & aids browsing experience.  User testers appreciated the amount of information listed per song, such as Energy, Length, and Genres as it made navigating through the songs easier.

Checkout form can be further enhanced. Participants felt that the checkout experience was “busy” and “cluttered”.

Mobile users face difficulty navigating filter options on the site. After interacting with the filter, it was still not apparent to a user that they could filter music by genre and they were not able to successfully do it.

Key Research Takeaways

After reviewing all of my findings from the Marmoset shopping experience, I wanted to address the opportunities I thought had the highest chance of increasing their conversion rate in my design recommendations:

- Filter placement and logic may encourage users to overfilter on desktop
- Navigating mobile filter options was cumbersome for users
- Sitewide radio buttons poorly guide users
- Mobile experience does not provide an easy way to play or pause a song.
- Users thought checkout felt cluttered

.

A/B Test Concepts

Based on my prioritized findings and feedback from my team, I highlighted 3 areas of the Marmoset site where I wanted to conduct testing site. The goals of these tests were to improve the online shopping experience for guests and conversion rates for Marmoset Music.I created high fidelity mockups to communicate my test concepts to Marmoset Music.

Original

My recommendation

Established Challenge: Functionality of radio buttons is not intuitive based on the icons. “Buy License” is located in the center, not being prioritized.

Testing Opportunity: Provide more directional guidance for users by implementing a visual hierarchy to highlight the most important actions to them, such as Buy, Download, and Favorite, and hiding secondary actions that are not used as frequently.

Goals: Increase interactions with "Favorites" + "Downloads" and transactions

Original

My recommendation

Established Challenge: Heatmaps and UserTesting indicate that users are not recognizing that the “Vocals” button opens up a drop down menu with more options. Overall, filters are not optimized for mobile users. 

Testing Opportunity: Reconfigure the mobile browse page to make it more intuitive by altering visual treatments of tabs and dropdown menu and pairing "Clear All Filters" Button with filters it impacts.

Goals: Increase dropdown menu interactions, song plays, and transactions

Original

My recommendation

Established Challenge: Prominent positioning of filter on the Browse pages encourages more interaction with this section, creating a tendency to over filter, resulting in users having to re-curate their selection.

Testing Opportunity: Highlight existing feedback surrounding what filters have been selected and the total amount of songs that fits existing criteria to make users more aware of how their selections are pairing down their options.

Goals: Increase song starts, transactions, and visits to checkout.

The Results

Unfortunately, due to the business impacts of COVID 19, my relationship with this client was cut short. However, my team presented this work to Marmoset Music on my behalf and although I did not see the breakdown of the test results, over time I have seen several of my recommendations implemented on the site. 

Reflection

I learned a lot about creating A/B tests from this project. I made test metrics, hypothesis, and goals. I recognized how the amount of online traffic can impact the amount of variants a test can have and how long it will take to have statistically significant results. 

I also saw the benefits of utlizing both qualitative and quantitative research methods to tell the story of the user with different types of data. For example, HeatMaps revealed that users were interacting with filter options less on mobile, and UserTesting revealed that users did not recognize this part of the interface as interactable. Pairing both of these metrics together makes for a painpoint thoroughly supported by data.

Carina Dempsey 2023