GEN OTSUBO
Sell-on-Amazon is a cross-platform smartphone app. This app makes selling items on Amazon.com easy and fun for small businesses and individuals. It offers simple and intuitive user experiences for keeping inventory and selling new/used items online.
I collaborated with an application engineer to create the app from the concept generation. Five business experts from the US joined the team in the later phase. My responsibilities in the team include:
This app evolved out from our web-based service for selling items on amazon.com. We redesigned it from the ground up for smartphones, using all of our knowledge and data from its previous projects. It was the first Amazon marketplace smartphone app in the world, which we later pitched to Amazon.com, Inc.
We initially targeted charity shops (thrift stores) in the UK. There are almost 10,000 charity shops in the country. They mainly sold donated books and clothes at low prices.
It was very common for the prices in the shop to be much less than the prices online. They sell faster online too. We saw a business opportunity and started making apps for sell items online.
We have talked to hundreds of charity shop managers and staffs to gain feedback about selling items on the Amazon marketplace.
Many of them were interested about the Amazon marketplace. Some even have tried to sell their goods. However, very few have started selling items on it or kept using the service.
Most shop managers were interested in selling items online for higher prices. They also wanted to use it as an inventory system to keep track of the item they own.
The complexity of the existing service was a high wall, especially for shops where elderly volunteers are working. Managers were also concerned about the online connections and the limited space at the till.
Making the app for smartphones solves both issues for limited spaces and internet receptions. We decided to move forward with prototyping the app.
I constructed prototypes to be tested by several shops. I created a minimum viable product made for books, with the following features.
Soon added a feature to estimate the time to sell, which the team has already programmed for the desktop versions.
I conducted online user tests limiting users to be above 65 years old. I set several tasks to add items, input information and answer customer questions.
Improved the UI design through these tests and feedback of users. Adjusted button positions and colours if the user did not find it immediately. Some users are not used to taking barcodes on the smartphone camera, so I tested the UI with variations of several supportive graphics and texts.
The key features of the application.
I preselected most options, including pricing and shipping details, to achieve a streamlined user experience. It works for 90%+ of all situations. We also added competitors' information for comparison, to quickly and accurately modify fields when needed.
1. Start by scanning a product barcode.
2. Product details are loaded: Check the price, and estimated time to sell.
3. Select a condition grade.
4. Hit the confirm button and the item is listed online.
Many features were irrelevant to small businesses and individuals. It is the main reason that made users give up before even trying.
1. Replenish Inventory
2. Create a new shipping plan
3. Choose items and quantities
4. Prepare products
5. Label products
6. Preview shipping
7. View Shipments
8. Prepare Shipment: Choose shipping service, Measure and weigh box, Calculate shipping charge, Print shipment slip, Print box label
9. Shipment summary
In our app, we have cut down on features that are not used by small-scale businesses. Refined interfaces, so all necessary features fit inside 4 smartphone screens.
1. Confirm items to ship.
2. Measure and weigh the box.
3. Confirm shipment details.
4. Two PDF files are emailed to you for printing.
In the original interface, sellers must navigate many pages to handle routine order-related inquiries.
Customer support is a very important feature for online marketplace. It is also known to be a greatest hidden cost. Small businesses must handle last-minute cancellations, or return requests almost everyday.
I paid particular attention to making customer support as simple and as intuitive as possible.
I used a familiar messenger style layout for customer service. We also added a customer timeline, where users can view all relevant information to reply to customers, and decide actions.
200 customer service cases are used to analyse the commonality and the priority of the actions.
The repetitive responses are automated and displaying choices only if seller input is necessary.
90%+ of information user needs to reply to customers is accessible in 0 to 1 tap.
For the US amazon, there are 20 domestic shipping services to choose from, presented in a single table. It is very confusing to sellers.
In our app, we made an algorithm to show the top 3 appropriate choices, which will arrive on time, cheap and reliable. In rare cases, the seller may tap for more options.
Gamification techniques such as instant feedback and scoreboards are used in places, to encourage sellers to ship orders and reply queries fast.
It was the first time working with people with over 30 years of business experience. I learned a lot from the business experts from the US, from effective presentational techniques to the best timing to call a manager for a pitch.
The ways users make mistakes sometimes even feel creative. I learned the importance of testing and improving the UI.
All design details affect users succeeding the tasks, from button positions to the amount of information on the screen. If I highlight one thing, the other information gets suppressed. I learned to test and iterate the design for the best balance of information and action triggers.
