In-Feed Language & Interest Selection
In-Feed Language & Interest Selection moved MX TakaTak’s language and interest preference prompts into the video feed. Instead of sending users to a separate setup flow, the product collected stronger personalisation signals through lightweight in-feed cards that felt connected to watching videos.
Results
- Language and interest choices gave the recommendation system clearer signals for content ranking and discovery.
- The flow encouraged preference submission without interrupting the core video-watching experience.
- Cleaner preference data improved content relevance, ad targeting, and monetisation opportunities.
Date
Team members
Cui Shanshan
Organisation
Feed-Native Prompt
The UI used the same full-screen feed frame users already understood: top navigation, creator avatar, bottom tabs, and a centered preference layer. That made the prompt feel like part of MX TakaTak instead of a detached settings task.
Language selection started with a disabled submit state, turned active after one or more choices were picked, and confirmed submission with a short success message. Users could then jump back into the feed without losing momentum.
Interest Selection
Interest selection followed the same interaction model, but supported a larger set of categories such as Scene Creation, Art & Crafts, Comedy, Movie Shows, Dance, Music, Food, Sport, Fashion & Beauty, Gaming, Technology, News, Travel, Health & Fitness, Devotion, DIY, and Motivational.
The flow also accounted for loading, selected, submitted, jump-to-next, and swipe-up states. Those details mattered because the prompt lived inside a moving feed: it needed to collect data, recover gracefully, and always point users back to the next video.
Personalisation Outcome
The design reduced friction around a high-value machine learning input. Users could express what they wanted in a few taps, while MX TakaTak gained clearer language and interest signals for recommendations, discovery, and advertising relevance.
The result was a concise, non-intrusive preference capture loop: ask in context, make selection obvious, confirm submission, and return the user to watching.
Ideas by me. Written by AI.
I’m explicit about how I write. The ideas, point of view, and responsibility are mine. AI helps with structure, clarity, and speed.
Why I work this way →Add a comment
Your comment will be sent to me privately before being published.
Like what you see?
Let’s team up
Together, let’s redefine your business’ product experience strategy.
You can contact me at: