Hard Problem Series: Next Gen Tiktok — Privacy, Usability, Engagement, Real World Benefit

A practical look at what our future social media experience could be

The banning of TikTok and other social media tools in India along with the negative connotations those platforms have as sentiment right now gives rise to opportunities. Plus it makes good practice to try and take the most popular apps in the world, learn from their UX to integrate it into a solution that solves the underlying issues inherent to the system.

Great product design and development will take a bunch of information and make some guesses in order to create a UX flow they can test to ensure the metrics they want to increase with this change actually increase.

First, it would be worthwhile to review some of the issues with these platforms in order to give direct requirements for further development.

What are the exact issues of these platforms:

• Privacy Concerns
• Engagement skews young
• Parents have no visibility or benefit of kids using the platform
• No additional value from reach on the platform
• Control by Chinese State
• Data hoard platform(Not unlike other social platforms)

What are the underlying general drivers:

• Increase new user conversion — Lower CAC
• Increase Engagement — Increase current user engagement or diversify user persona engagement?

User Persona:

Younger 6- 13 — Kids who want to make fun and entertain content that involves themselves or just as a content intake flow.

Parents — 28–50 — Parents who are slowly letting their teens use social media, but want privacy by default and to monitor their kids.

Use Cases:

1. Sign up — Content intake is a simpler flow, so we will focus on the content generation for now. In future, we may look at this in more detail to determine some small improvements that might drive value here.
2. Video Creation — Kids or Parents see a video they want to make.

Quick Research:

For our purposes, I am not going to show all the work required for this phase. Essentially we will need to do an inventory of the competitor flows, and understand better how we will design it based on our use. Much more research would usually be done in order to get deep into the details of that the user flows. For instance, below we will review the UI and UX of Tiktok and Instagram.

Similarities:

1. Video Playing Full Screen
2. Follow Button or Profile Button
3. Share Buttons

Tiktok Increased Engagement Items:

1. Love or like button visible
2. Comment button visible
3. Duet
4. Music button and info visible

The hypothesis to reach requirements:

We will be bringing a number of different industries and fields in order to take what we hope will be the best approach from a UX perspective in order to reach the goals provided. Below we will go through some of these ideas/notions in order to develop some UX proposals.

1. Games are engulfing the world over videos — Pokemon for instance has brought in 17 Billion. Movies/Videos seem visceral.
2. Games as a service — To keep revenues and engagement high game developers are starting to turn to see it as a game as a service they update regularly. Fortnite for instance has new game features and weapons weekly.
3. Real Word Engagement — Such as with Pokemon, and VR coming of age is showing a trend of moving the digital experiences to the real world.
4. Shame — We have to ask why someone wouldn’t engage to create content. A few reasons come to mind. 1. Shame — just don’t want to be in a video, 2. have no reason or friends to make the video.
5. Fun — As we are targeting younger people and many apps like Tiktok show us how making the app more basically fun can increase engagement and usage. The question will remain what unique feature we can add in order to make this still true.
6. Privacy — No user logged in required to watch or record.

Let’s define and build something:

Now that we have reviewed some important trends we can put together a simple list of high-level features.

Pitch: Move to the beat of your own body.

Features:

1. Privacy first — No need to register to watch or make videos.
2. Battle — Challenge people to match dances and compete to get the best dance scores based on a posenet and facemesh sentiment model.
3. Mask me — Overlay robots or other characters over the dancer that represents the user's body and the video generated can use this 3D model or not.
4. Trainer — Use the scoring system and overlays to help train users in different body movement exercises. This could be beneficial for dance, karate, exercise, yoga, and a number of other areas.

Technologies:

What technologies and platforms will help us the best while adding the additional features we need in the easiest way.

1. First, where can we find the battle and Trainer technology for scoring? We can most likely use this for the mask my feature in the future as well. https://www.tensorflow.org/lite/models/pose_estimation/overview
2. Javascript language
3. Multiplatform build: PWA React

Examples:

For this simple app we will focus on understanding the main features and how they function. IE mainly a simple app that can take dance and turn it into a game. For this, I focused on YMCA and still images as a starting point to understand the technology. This will help me understand how the comparison of the pose is computed against static examples compared to videos in the future.

1. Github Repo: https://github.com/Deamoner/dance.studio
2. Demo Site: https://posenet-ymca.herokuapp.com/