Why did I create a startup?

My goal was to understand how to run a product from more than just a design perspective

Working at a large cooperation, there are always many talented experts from different fields contributing their own expertise to the products. This can sometimes lead to the scope of the project being lost. I believe understanding all stakeholders’ goals and considerations upon digital products will help me to make design decisions faster and more precisely. Thus, I decided to create a mobile app from scratch to immerse myself in the whole process.

Design

TChat took place in two design phases: MVP and V2.

MVP version (minimum viable product) was created to push TChat out to market as soon as possible. It only had fundamental features to allow users to start trying TChat; meanwhile, we started to improve TChat’s features according to users’ feedback.

V2 version includes new features based on all users’ feedback and improved the workflow and user interface design.

MVP version

Product Requirements
- Users can comprehend how TChat works
- Assign tasks to themselves and their friends and family
- Users receive notifications when a task is due
- Educate the security policies and protect users’ data
- MVP must be completed in time

↓

Essential Functions
- Account: Google and Apple Sign Up/Login
- Onboarding: Tutorial, and Notification Access
- Task: Title, Description, Assign to others, Set up Due, Leave Comments
- People: Friend list, Add by Email/QR code

↓

Adopting Flat Navigation

The two important tasks on TChat are tracking ongoing tasks and creating new tasks for others. It’s very common for old tasks to trigger users to create a new task, like the waterfall method. Compared to other types of navigation, flat navigation can provide users a quick way to jump between the task list and friend list to create a new task while browsing ongoing tasks.

Essential Functions
To provide adequate features for users to finish their goals

Information Architecture

Level 1
Since TChat doesn’t require very much personal information from users, using Google / Apple Sign up and Login is the most efficient way to let users use TChat. It also reduces the time for developers to implement account-related features.

Level 2
The accesses of two main functions, “Task” and “People” are provided in the bottom navigation bar. It allows users to quickly switch between these two categories.

Level 3
Users can find these functions under Level 2 categories which can assist users to complete the actions of using TChat. To decrease the commitment of tracking tasks which the users spend 90% of their time on, I decided to not provide access to the settings page in order for users to focus more on task-related missions. Instead, users can find it on the friend page.

Workflows
Two ways to assign tasks to friends

Task Page’s Header
Tracking ongoing tasks is one of the most frequent actions on TChat. The design should help users quickly locate the tasks they are looking for. Since people can assign each other tasks, the task list should be separated into two sections, “my to-do tasks” and the “tasks I assigned to others”.

The header should be bold enough to allow users to quickly identify my to-do tasks and tasks I assigned to others because the UI of the task list looks very similar. Instead of using the Segmented Controls from iOS or the Tabs from Material Design, TChat uses a uniquely bold and glanceable header design. Also, users can jump between these two sections by swiping left and right.

Task Lists

Users’ actions on “My task” and “Their task” are different. People go to the “My task” section to see what task needs to be done by them and then mark the task as “done” when finished. For the “Their task” section, people can check the status of tasks they’ve assigned to others, but often the assigner doesn’t mark “done” themselves. Based on these differences, I listed these required UI components for the designs.

My task - Radio button, task title, assigner, and due date
Their task - Task title, assignee, and due date

Add Friend

In the first version of TChat, the only access to add friends is an icon, but many age 60+ testers didn’t understand the “add friend” icon. Thus, we replaced the add friend icon with text. This change allowed all users to recognize the “add friend” feature.

TChat provides two ways to add friends because users have a plethora of reasons to download and use TChat. For example, a family may decide to adopt TChat to deal with their daily tasks at home. In this case, scanning a QR code on each other’s devices is the fastest and easiest way. On the contrary, scanning QR codes is not as useful for people not near each other. For advanced users, TChat provides another quick access QR code scanner on the right side of the friend search bar. This feature eliminates an extra tap making it even faster.

Create Task
The create task button adopts material design FAB (floating action button) located on the bottom right corner because the “Their task” button is located on the top right corner. Also, the bottom right corner is easier for users to tap when using one hand.

After tapping the button, the “Create task" sheet will slide up and fill half the screen. It contains the task title, assignee, task due, and description. The “Task title” must be filled in. In this case, the default assignee is yourself without a due date.

Task Detail View

After tapping any task, the task detail view will show up. In case users accidentally put in the wrong information, they can modify it in the task detail view. Also, there is a comment area for users to ask for more task details or for an update on the latest task status. Of course, users can delete the task if they don’t want it on the list anymore. TChat will notify users if any changes have been made.

New Skills, New Certificates

I have always been a full-time designer, so there are many things out of my scope. To fulfill this product, I had to get more familiar with app-building platforms, such as Google Analytics, Admob, Firebase, Google Ads Manager...etc. During the year I worked on TChat, I received more certificates than in my entire life. It also opened my mind to understanding how digital products run.

Firebase: A platform for creating mobile and web applications without building its own server.
Google Analytics: A tool to learn users’ behaviors on Apps and/or websites
AdMob: An channel for App owners to connect their mobile app with ad providers.
Ad Manager: Managing all ad spaces across different platforms or products.

What I learned

After learning these new tools, I discovered many interesting insights to help me adjust my design decisions.

a. From Google Analytics data, 94% of tasks don’t have task descriptions
In the first version of the TChat design, I thought the task title didn’t have enough for users to explain a task, so I added a description space for users to add more details about the task. Based on the GA results, I removed the description space from the task-creating page to simplify the steps of task-creating. However, to keep the other 6% of the tasks functioning, users could still add descriptions on the task detail page.

b. Users didn’t want to see ads while still learning how to use the app
One of the ads classes I took explained how important user experience is when they open an app for the first time. Although the ads’ revenue is crucial for TChat to survive, allowing users to successfully learn how to use the app and adopt this tool to assist their daily life is even more important. In the long term, it can create more revenue. Therefore, we decided to update the app to only show ads after users have been logged in for 5 days or more. As a result, the retention rate increased by 22%.

Marketing Strategy

After the product is rolled out to market, marketing plays an important role in reaching more potential users. TChat spent about $200 monthly on Facebook Ads. To spend ads credits wisely, I did many experiments to try to get as many new sign-ups as possible within this budget.

For example, we targeted low eCPM (Effective cost per thousand impressions) countries to reach as many people as possible due to low marketing budget. Then, I narrowed down the target audience by the following steps:

1. Pick high performing country according to CTR (Link Click-Through Rate)

2. View ad performing data and then target high performing gender and age groups

3. If the CTR number is raised, run three different sets of ads to test ad contents.

4. When high performance is found, enlarge the age and gender group again to test new ad content to see if it attracts a new potential group.

5. Repeat for the next campaign.

What I learned

This method I created worked pretty well. The number of new sign-up users significantly increased after the successful experiments. However, while applying this method, I found a few ways to improve:

a. Low eCMP country also has low CPM revenue from ad units
For ad support ads Apps that have enough budgets should focus on which countries have more potential for using the App. This factor is way more important than the number of people are reached.

b. Avoid advertising on countries’ holidays
Digital advertising systems like Facebook, charge advertisers by how much money you bid for showing your ads to the group of people on specific dates. Most well-known daily goods brands, like Coca-Cola or H&M, usually invest tons of money to boost their sales during the holiday because people tend to shop for goods during holidays, as a result, the cost of CPM increases insanely. On the contrary, most Apps are not related to holidays, and people normally don’t increase what they download during this time. Unless the CMP cost is set up manually, posting advertisements during holidays is an unnecessary waste of resources.

c. The first two seconds of an ad is make or break
By studying our ads performance data, I found out that if the audience watches the ads for more than two seconds, they have a very high chance to convert (click the call to action). A couple of factors can tempt people to continually watch the ads, such as attractive headlines, emotions, a sense of curiosity, and pertinent promotion info. TikTok is a great example to explain the behavior of those browsing. After understanding this, I always imagine I am creating content for TikTok which needs to sell a story extremely quickly.