Timeline: 1 week
Role: Understanding the friction area and designing 2 new features
Marketplace is a product in the Facebook ecosystem that evolved from humble "buy & sell" groups in facebook. It is a peer to peer selling and buying platform within the facebook community primarily aimed at selling used products. Lately, facebook started letting businesses venture into the P2P domain making it also a B2C platform.
The navigation is not personalized to the needs of the users. The buying and selling experience is not efficient when it is comes to Multiple listings.
Improving the Information Architecture of the navigation and customising it to the needs of the user. Using price tagging akin to Instagram's when multiple items are sold in a single listing.
Consumers who are mostly looking for second hand goods that are cheap. They know exactly what they want to buy and search for those products. They usually trust the buy & sell groups they are a part of.
People who sell second hand goods when they don't need them. Usually selling it to people from their community. They aren't really worried about the money they make through these transactions. They want an easy way to sell their products with minimal interaction.
People who sell upcycled goods for a slightly higher price to make money. They are more worried about their ratings and how they come across to their buyers.
My main focus is on Buyers and Individual/hobbyist sellers as they form the majority of the ecosystem at facebook marketplace.
I found 27 issues which focused on error prevention, consistency, recognition, flexibility and efficiency.
2 pages stood out that had the maximum number of Heuristics violated. These were the pages I focused to probe on in my Qualitative interviews:
1. Home Page (Navigation)
2. Product Page
A coded spreadsheet with all Heuristics violations in the application based on NN's 10 Heuristics principles
Qualitative User Interviews
To understand my problem space better I interviewed 5 people of mixed demographics who are active users of facebook marketplace. This gave me an in depth understanding of the users pain points.
User Interviews conducted on Zoom
Some of the findings through qualitative interviews made my problem definition more solid. I found two problems to chase
1. Making personalized categorization and
2. Handling listings that are inconsistent.
My scope involved tackling Generic Categorization and Inconsistent Listings, as these were immediate concerns voiced by my users
It seemed like the navigation was lacking clarity, here you can see how the actions are mixed in with categories. These are completely different from categories like "Local" "Shops" "Live Shopping".
Another problem area was during multiple shopping. When a person is selling many items in a single post they provide pricing for each item in the description. This creates mental load on the users.
Usability Issues uncovered in Navigation
Usability Issues uncovered in Multiple listings
I created personas to focus my redesign and understand the motivations, problem and goal of my users. My target user groups are student buyers looking for cheap/ second hand items and individual sellers who want to get rid of their items.
1. Generic Categorization
Most of the users felt the categories that were shown in the top navigation of marketplace were impersonal and did not reflect their buying needs.
“These categories are not even remotely close to my interests”
2. Inconsistent Listings
Multiple items being sold in a single listing is hard to keep up with for buyers.
“Usually the listings are marked a different rate because there are multiple items being sold in a single post.”
When it comes to buying any item, the users usually check if they are available in their buy sell groups after which they move to the marketplace.
“I usually buy things from buy/sell groups I am part of because they are a trusted network of people.”
4. Seller location accuracy
Users without vehicles of their own worry about the location of the seller so they can understand the logistics. Currently facebook provides a larger location range which could be made more specific.
“I need to know where the seller is because I don’t have a car and have to think about transportation"
Problem 1: Tackling Generic Categorization through improved browsing experience and personalization.
1. Improved Browsing Experience
Open Card Sorting:
To make a more data driven, customer centric decision, I conducted an online open card sorting experiment on Optimal Workshop to understand how to make navigation and browsing better.
No. of Cards: 30
No. of Participants: 10
Total time taken to complete: 12 mins on an average
A dendrogram of how people categorized the items provided in facebook marketplace
1. Most of the people placed sell search and profile in the same group "Account" "Actions" or "Stalk"
2. People had an average of 7 buckets which was interesting and could be further understood but it was outside of my scope right now
3. People also grouped Local listings, Live Shopping, Buy and sell groups and deals together and not under categories.
2. Personalization of categories:
I went ahead to understand how other products tackle the problem of content personalization. But also give the flexibility to explore categories if they want giving power and flexibility to the users. I later studied how the selected content is displayed on screens. Medium does a great job with categories selection by writing a small gist about the category. For content display, Pinterest uses "For you" which I thought aligned with my design.
Medium and Pinterest do a great job in personalizing the experience for their users
Problem 2: Solving inconsistent multiple item listings through price tagging of items on images.
Price Tagging of Images
I took instagram as an inspiration as they do a great job adding price tags on items that are sponsored by influencers. Since instagram is a subsidiary of facebook, I followed a similar design style but would have to design the flows which would be unique to the marketplace
Studying price tagging interaction on instagram
I iterated a few flows that I thought made the most sense and finally came up with two interaction flows. I chose to go ahead with option 2 as it had fewer screens to get to the final stage.
1. The images are more focused.
2. Flow is more intuitive.
1.Must get through more screens to finish the task
2. Reduced redundancy
1. Getting to know prices is quicker
2. Reduced steps (only 4 screens to the finish)
3. Follows facebook's design language
1.Images are not the hero
Navigation and Personalization
Multiple item listing
Improved navigation & Customization
The sell & search actions are made more prominent as they are the most important features for both Rosa (Seller) & David (buyer)
David can add his favorite categories that are relevant to his interests without letting go of the exploratory aspect of the Marketplace's Home page.
Price tagging multiple items on an image
Rosa can now price and sell all her items in a group with ease, without creating multiple listings.
Price tagging multiple items on an image
Improved buying experience for David for listings with multiple items.
The price is shown as a range rather than a randomized number for better decision making.
Negotiating with the seller is easier as the seller knows exactly what item is being asked.
Seller's flow when using Marketplace to sell multiple products
The problem is solved by using price tagging, where the price can be mentioned ON the image without using the description box to list down the prices of every item.
Buyers flow when they come across Multiple item listing
Buyers can now easily navigate through the listings and communicate effectively with the seller on listings with multiple items without having to send screenshots.
Defining the problem areas before interviews made it possible to dig deeper
Since the time frame for the redesign was only a week, defining the problems areas through heuristic evaluation made it possible to screen the participants and drill them down further in those areas. Avoiding exploratory research saved me time which I used to conduct early usability tests.
Always drive your design using data
Although I had a basic idea of how my navigation needed to look. I went ahead and conducted a card sorting experiment to make a more data driven decision. This helped me validate my design decisions and helped me build a more customer centric flows.