Smarter Listings 2
The second part of the Smart listing case study focuses on how we integrated Smart listings with the buyer experience. If you haven't read the first part, I highly recommend going through it!
Before I move forward, if you stumbled up on this article directly, some context
Carousell is a C2C Marketplace app expanding from South East Asia. The product focuses on casual sellers, like people with a day job or school to de-clutter their space by start selling.
Who I worked with?
I closely worked with Ari, Senior Product Manager, Andrius, The VP of Product and the Smart Listing, Vertical Buyer Scrum teams to launch Smart Listings for buyers
Platforms & Counrties
How Long did it take?
The second phase took 5 Sprints to design the buyer experience with the smart listings
Reiterate on the problem we started with, we wanted to build powerful carousell listings that would make us future proof in terms of buyer/seller experience and also from the business point of the view
We solved one half of the problem on the supply side and we enabled listings with comprehensive information. So now,
Current Buyer Experience:
The goal of our team is to not change the buyer experience flow significantly, but just add enhancement to the experience. The First step is to figure out where are the crucial parts of the flow we need to enhance the experience.
The summary of the above journey: A Buyer with their intent to buy an item discovers a listing and spend some time reading the details in the listing to decide to buy or not. This is a compact version of the journey and has been positioned to highlight the parts where Smart listings can enhance the experience.
We identified two crucial problems that are hampering the browse and
1. How might we help buyers to refine the results further?
2. How might we showcase our listings better, so it helps our buyers to browse?
These two questions directly touch up on few areas of the buyer experience. We decided to focus on the highlighted areas as there will be the maximum impact.
Product Cards in
Search and Browse
Chats & Transaction
1. How might we help buyers to refine further?
In the previous article, we talked about various details sellers can now provide for their listings to make it rich. We now want to enable our buyers to refine the results when they are browsing. Our hypthesis is that refining results will yield more relevant results, and thereby higher conversions.
Current Sort & Filters
When a buyer searches for an item or browses in the marketplace, they can refine and see relevant results faster by Filters and sort.
Our current Sort and Filters are simple by its nature, echoing the simpler listings in the marketplace. Buyers can sort the listings by popularity, recency, price, and how near the item is. Buyers filter the listings with price columns, item condition and how can they get the item - meetup or mail delivery
While our current sort and filters served it's purpose for simpler goods, when we are scaling to properties, cars, jobs and services, we want buyers to be able to sort and filter results with the attributes that make sense for the listings they are browsing in Carousell.
Like our new dynamic smart sell flow, we wanted to build a filters flow which is dynamic and can be built to cater different categories.
Starting with the Sort
We enabled the dynamic sort for different categories. As a result, the Sorts in the jobs category will be different than sorts in Women's fashion. The design also extends in adding any new category and new sort along with it in one go.
With Filters, we wanted to enable different possibilities in letting the buyers choose what they want. Filters come in all shape and size. Some are direct and lets the user input the preferences directly through checkbox, radio, text/number input components, while some of them require users to select certain fields from the lists.
Showcasing some examples from different categories.
Things we couldn't do
For various reasons like Scope, time constraints, dependencies, and lack of research, we couldn't do a lot of changes in the buyer experience as we originally intended to do.
We couldn't improve the discoverability of filters in the apps. While we know it's an issue, we wanted to have sharp focus in delivering what we wanted to do and decided to ultimately pass this problem to Buyer Experience Team
- While we improved the Filter diversity & utility and gave more control to the user, the experience still not as compact or not tailored for certain use cases - Like a better size picker or picker that has icons/colours would have boosted the usability
- We couldn't validate the filter experience properly because of the time constraints. As a result, I was only able to internal user testing.
2. How might we showcase our listings better?
Currently, listings are presented in terms of product Cards. Product cards are compact forms of listings and they showcase photos, and few crucial information like title, price and seller's information on the card.
This worked very well for our old listing type because of the simple listing type that featured title, price, user information and photo. But when we started to expand into property, car and job listings, we wanted to show additional information that's useful when the buyers/seekers are browsing through the listings
Listing card Structure:
We wanted to show two additional attributes that showcase the value of the listing while browsing. But adding them just below the price and increasing the card size would make the card look long. We rethought and re-organised the card information and increased the card height by 15% and the new product tiles support 5 attributes to show various information. Additionally, we also designed to enable the dynamic call to action in the cards, more useful in Profile page to promote or list their item.
Showcasing how it would work for a Women's fashion listing in marketplace and profile.
Things we couldn't do
- We wanted to test out the product cards and see how it performs. So we tested it in browsing and search experience. After the test, due to the time constraints, we couldn't scale the product cards to few more important pages in the app. As a result, two cards exist in the app right now
- I didn't design for the future, something I am trying to overcome as a product designer. As a result, the product cards didn't age well in the product. When we launched the nation square photos for the products, the listings show the square cropped photos and it's not the good experience.
When we first started to work on the selling experience, we know the solution we come up with goes hand in hand with the listing details experience. In short, we asked sellers few mandatory questions for each kind of listing that buyers likely want to know. We wanted the new listing details page to extend the current listing experience but at the same time highlight the details of the item that is more important.
The current listing page tells a story about the item and the seller. While buyers can identify the item by looking at the title, price and the category it belongs to, description is usually the part that tells a detailed story. Apart from that, there are few more components highlighting the quality of the item, how can they meet up or get it delivered and more details about the seller.
One thing we learnt from working with the seller experience is that designing components helps us a lot and easier to scale to various categories.
This mental model helped in coming up listing details components. Because of the nature of the page, we also had to rebuild the existing components.
We also designed these components in a way that order can be defined by us, so we can give a unique & contextual experience for different categories. The components are also flexible in a way that based on what information we want to show, we can position it with/without an icon
Things we couldn't do
1. While we built the product that can scale to different categories, the customisation took a lot of time and our team couldn't scale for all categories. As a result, even iterations on top of the customisations cost us a lot of time
2. One of the feedback we got was that the page becomes long as we tried to show more information. The other problem is also how we show the info. The design is to blame part of it as we took a safe bet of having single column layout rather than trying out with two column one.
3. User test the listing details page to understand the pitfalls. While we tried to keep the experience as close to the previous listing details page, the problems also followed through. We should have made the page more accessible, by supporting dynamic font sizes in the phone and giving the way for users to glance through the information faster.
All the parts of the buyer experience improvements covered in this article were tested against the previous experience to check if they actually add value to the users. Overall the filter experience and listing details experience didn't affect our health metrics except for the product cards.
One of the best ways to see how users use the app is to communicate regularly with the customer experience team and see app reviews. We found few issues in usability and issues in how we are showing information. We are currently handling how we solve these issues through Vertical Buyer team and extend to the entire marketplace.
Case studies from my work & fun Projects