Social media platforms often struggle to balance algorithmic recommendations with organic discovery, leaving users with either overly broad or overly narrow content feeds. Twitter's current systems—like the "For You" timeline and trending topics—don’t fully leverage the platform's unique strength: its dense, interest-based social networks. Many users miss high-quality tweets from accounts their mutuals follow or topics their network engages with, simply because they aren’t directly connected to those sources.
One way to address this gap could be to introduce a dedicated "Explore" tab that blends social proximity with interest-based recommendations. Instead of relying solely on algorithms or broad trends, this feature could surface tweets from accounts followed by a user’s mutuals, prioritized by how often those mutuals interact with them. For example, if several people you follow engage with a particular account, its tweets might appear in your Explore feed. This approach could be supplemented with subsections like:
To avoid creating echo chambers, the tab could also include a small percentage of serendipitous content from outside a user’s immediate network.
This approach would cater to several groups:
For the platform, this could increase engagement by making discovery feel more organic and less forced. Advertisers might also benefit from richer interest graphs based on both user behavior and network activity.
A minimal version could start by showing tweets from mutuals’ follows, ranked by engagement metrics (likes, retweets) from those mutuals. Over time, additional layers—like the subsections mentioned earlier—could be added based on user feedback. A/B testing could help balance social signals with algorithmic recommendations, ensuring the feed feels personalized but not insular.
While platforms like Instagram and TikTok offer discovery features, Twitter’s version could stand out by leveraging its text-based interactions and tighter social graphs. The key would be to make discovery feel both relevant and serendipitous—a mix that’s currently missing.
Hours To Execute (basic)
Hours to Execute (full)
Estd No of Collaborators
Financial Potential
Impact Breadth
Impact Depth
Impact Positivity
Impact Duration
Uniqueness
Implementability
Plausibility
Replicability
Market Timing
Project Type
Digital Product