Tmall for Business (Project at Alibaba)
An internal tool that helps the marketing operations specialists filter qualified products for marketing campaigns to achieve business objectives.
What is Tmall?
Tmall is one of the largest B2C online retail platforms in the world, operating under Alibaba Group. Tmall features more than 70,000 international and Chinese brands from more than 50,000 merchants.
What is the problem?
The existing process of filtering qualified products for marketing campaigns is inefficient and inaccurate.
What is the goal?
The design goal is to build a new web-based data visualization system to help the marketing operations specialists filter the qualified products easier and faster so that they can exceed the business objectives, which is to increase the click-through rate, conversion rate, and GMV.
What did I do?
As the only Interaction Design intern on the Tmall UED team, I have been able to conduct user research, develop wireframes and prototypes by using Sketch and Principle, and perform user testing to iterate my design. In the end, I worked with PM and developers to ship this product.
Why does a marketing operations specialist need to filter products?
Before designing the new system, I had to get a better sense of its current state. So I started the project by meeting with product and operation teams and thoroughly explored the business. These efforts helped me establish a rough understanding of why we need a new system.
How does a marketing operations specialist filter qualified products?
In order to know the existing filtering process, I conducted 14 user interviews with our operations staff, product managers, and front-end developers. Since the primary users are marketing operations specialists, I also observed how marketing operations specialists collect and analyze data.
To better guide our design and enable everyone on the team to empathize with our users, I further synthesized the research results and came up with the key persona with her journey map.
Due to heavy workload, users only collect data from pre-selected suppliers, which limits the opportunity of other suppliers
In the fast-paced environment, users need to comprehend data quicky
Users evaluate data based on their experience, so the filter list they create may not be accurate, which may lead to inaccurate results.
There's no way to test if the filtered products are qualified for marketing campaigns.
It’s hard to store and track results due to lots of Excel sheets
While this project aims to build an MVP, it is both unrealistic and not very user-friendly to build all the functionalities without discretion to meet all user's needs. So I worked with PM using the MoSCoW method to prioritize the criteria based on the insights and business needs to help me identify the most critical functionalities to design and implement.
How might we design a filtering tool that
improves efficiency and ensures accuracy?
With all the problems, goals, and requirement in mind, I started to ideate some possible solutions matched against each of the requirements and then discussed with the team about the proposed new user flow.
New User Flow
The next step is to build a prototype. Working with PM, I started by collecting and analyzing the types of data and filters that need to be represented in the system so that I can organize and structure all relevant entities in the information architecture.
After created the information architecture, I sketched a low-fi wireframe to show how it works. After several rounds of discussion with the team, I iterated the low-fi wireframe as shown below.
Home Page - Merchandise Pools
Filter & Data Analysis Page
The List of Filtered Merchandise
Users can create a new pool by a second round of data filtering to ensure the accuracy
choose a pool to view the data analysis
The List of My Pools
Q: How to design 100+ filters?
A: Using data as a filtering tool.
To find out if the concept works, I brought the design in for evaluation and testing with the stakeholders. The key feedback from operation teams are:
1. The new tool improves efficiency, even though there's a learning cost.
2. Functionality over aesthetics for internal tool/back-end product.
3. Information overload: displaying all data diagrams in one module makes the site too long to read and reduce efficiency.
Based on the feedback, I made several rounds of iteration, tweaking some details to make the flow more intuitive and useful.
For the reason of confidentiality, all data referenced in prototypes are made up.
Filter & Data Analysis Page
The List of Filtered Merchandises
The List of My Merchandise Pools
Know the business
In school, I thought that designers were expected to focus on how things look and work. The real-world experience taught me that design can deliver business goals. I am glad that I used to study business because it helps me quickly understand the business background I started the design. However, how can design thinking be better integrated with business practices to improve outcomes for enterprises is the most challenging part of my internship.
Listen and communicate
It happens a lot when our colleagues, such as developers and product managers, have different opinions with designers. So being about to listen actively to others and tell a powerful story about our design will make the collaboration deeper and easier.
Learn HOW TO LEARN is important