Prediction Economy

The underlying assumption of the prediction economy is that user behavior is somewhat modelable and can be estimated in advance.

Under the prediction economy, the substantive product platforms sell shifts from space and time to the ability to predict the conversion, churn and upgrade of specific users.

Predictive capability can feed back into content operations, judging in advance the effect of subject matter, timing and audience based on historical data, forming a barrier that is hard to replicate in the short term.

The feedback loop is the key mechanism by which the prediction economy keeps running; a prediction system lacking a closed loop will fail over time due to drift.

The threshold of the prediction economy is higher than that of the attention economy, requiring long-term accumulation of data density, modeling capability and a feedback loop, and may widen the capability gap between leading platforms and small and medium-sized media.

Under the prediction economy, the metric system by which advertisers measure campaigns shifts from impressions and clicks to the estimation accuracy and actual hit rate of conversion probabilities.

The prediction economy faces ethical issues such as the boundaries of data collection, users' aversion to prediction, and the risk of prediction being used to manipulate rather than serve.

The media industry is shifting from the attention economy to the prediction economy, and the industry's value anchor is shifting from impressions to accurately predicting user behavior.