Coins, cliffhangers and the maths of episode four

Every short drama app is solving the same problem: find the exact moment a viewer is invested enough to pay and not yet satisfied enough to stop.

That moment is usually the end of episode three or four. It is not a creative decision. It is the point where the free window has done its work.

The three models

Coins. The viewer buys a currency, then spends it episode by episode. Revenue per paying user is the highest of the three, because the price is disguised and the increments are small. Conversion is the lowest, because the total cost only becomes visible after several purchases.

Ads. Unlocks are earned by watching. Revenue per user is low and the experience is worse, but almost everyone converts. This is the model that works in markets where card payment is friction.

Subscription. Flat monthly access. The most predictable revenue and the easiest to explain, but it caps the ceiling: a heavy viewer and a light viewer pay the same.

Why AI changes the floor, not the logic

AI production lowers the cost of making episodes, which changes how many episodes a season can afford to have. It does not change where the paywall goes, because the paywall is placed against viewer psychology, not production cost.

What it does change is the tolerance for a weak episode. When an episode costs a fraction of what it used to, there is less excuse for padding the middle — and viewers who arrived for the hook notice padding faster than they notice a softer image.

The apps that rank highest on our AI drama apps ranking are not the ones with the most aggressive paywalls. They are the ones where the free window and the paywall are placed with some care for what the viewer has actually received.