top of page

E-Commerce
Generative Study

Project Summary

I compared the sales funnel efficiencies for a client working in e-commerce. The study shows the impact on views, likes, and sales revolved around the title formula. I launched a classic A/B test on 4 different items being sold at the same time, on the same platform. All variables between the 4 items were the same (price, pictures, item descriptions, shipping costs, etc.) The only variant between A and B of each item was the listing title.

Note:

To protect my client's privacy, the actual formula used by the client is not shared in this report. 

Project background

 

How it started:

​

  • What user problem are we looking to solve?

    • The client has been selling on the resale platforms Mercari and eBay for 2 years. At the beginning of her career, Mercari had a shorter character limit than eBay for listing titles.

    • In recent months, Mercari extended the character limit from 40 characters to 80.

    • She implements a cross-listing strategy on multiple platforms for the same item and observed her items on eBay were selling faster and more frequently than Mercari. She wanted to know if the same listing title formula used on eBay would prove as efficient on Mercari.

​​

  • Stakeholders:

    • Client: Susan

​​

  • Research goal:

    • Primary goal:

      • I needed to design a research study that would definitively show the impact of implementing eBay's title formula on Mercari​

​​

My Role

 

  • UX Researcher

 

Method

 

  • What are the methods I chose?

    • In-Depth Interview

    • A/B Testing

​​

  • Why did I choose these methods instead of the others?

    • I was able to meet with my client in person for a couple of hours which gave me a chance to go in-depth on what her challenges were and gain meaningful context around the e-commerce strategies she uses and learn what her successes have been over the years.

 

Challenges and Limitations

 

  • What challenges did I face during this research project?​

    • Visibility in the buyer pool is very limited

    • The user that I'm solving for is the seller - not the buyer

​ 

  • What were the limitations on budget and timeline?

    • The test ran for 14 days​

    • We purchased 4 pairs of items for the sake of the test, which did require some investment

 

Crucial Findings

 

  • The longer titles gained the most views

​​

  • Likes and Sales had very little impact

 

Implementation:

​

  • What was achieved?

    • Susan learned that eBay's listing title strategy works across other selling platforms as well.

    • She observed that Mercari's search algorithms can be manipulated to help buyers find her items sooner than before.

​​

Stakeholder Impact

 

  • How did I collaborate with stakeholders?

    • While designing the study, we narrowed down 4 separate pairs that would be listed individually (8 total listings), while keeping overhead investment minimal

    • We kept the test time frame to the average time-in-market as her other listings

​​

  • What KPIs were impacted?

    • Views

    • Sales

 

Follow-up from this project:

  • The next test Susan would like to run is to see if hashtags speed up sales on items listed on Mercari

 

My learnings

 

  • If I did this again, how would I do this differently?

    • I would attempt to learn more about the buyer pool with some audience tracking of some kind. Perhaps the suite of listings my client has to choose from could be more optimized for a specific audience who uses Mercari vs. eBay.

bottom of page