Korva is a scroll of the internet’s best ideas.
Keep 10% of creators exploring
Running the winning format
Moving most of the cohort onto the winner works because every creator has seen another creator in the cohort do well. They know the format works, and with a view bonus on the table they put extra effort into copying it and dragging it out as long as it lasts.
Why some keep exploring
The one or two creators still testing matter for two reasons: new formats keep turning up, and a backup is ready to slot into the chamber when the main format gets too saturated. Some teams run a single format into oblivion. A slightly smaller basket on the side hedges against wearing out the audience.
The first videos of a new breakout format are often the ones that go the most viral, because the algorithm hasn't seen them before. Creators in the discovery phase get the best shot at those.
Adding new creators
Adding creators gets easier once a campaign is winning. Creators are more willing to join when they see it's working and believe they can become one of the top performers. A five to ten minute call covers how the campaign is going and why it is scaling, and then the new creator is assigned to the campaign and added to the group chat.
Why do the first videos of a format go viral?
A recommendation system shows a new video to a small test audience and widens its reach as viewers keep watching. Early videos in a format face little competition from lookalikes, so viewers have no sense of having seen it before. As more accounts post the same format, each new video feels more familiar and holds attention less. TikTok has described its system as expanding a video's audience step by step based on how earlier viewers respond.
Why does seeing a peer succeed motivate creators?
Psychologist Albert Bandura argued in his social learning work of the 1970s that watching someone similar succeed raises a person's belief in their own ability. He called the belief self-efficacy, and observing a peer is one of its main sources. A creator who sees another creator in the same cohort do well has proof the task is achievable. A view bonus adds a payoff to that belief, so effort goes into copying the format.
Why keep exploring when one format is winning?
The tension between using what works and trying something new is known as the explore-exploit tradeoff. It is studied through the multi-armed bandit problem, named after slot machines, where a player must decide how often to pull a lever that has paid out against testing others. Statisticians trace the problem to a 1933 paper by William R. Thompson. A small share of effort kept on exploration is a common solution, since the best option can wear out or be overtaken.
More from The Superwall Podcast
