Why timing a post matters more than its content

An item is published and its first minutes determine how widely it is shown. Those minutes contain whoever happened to be online, and their reaction is used as evidence about the item. Content still matters, but it is evaluated by a small early sample whose composition is set by the clock.

What everyone sees

Reach is read as a verdict on quality, so a poorly performing item is assumed to have been weak. This treats the measurement as independent of the moment. It is not: the ranking signal is generated by early viewers, and which viewers those are was decided before anyone read a word.

What is actually happening

Muchnik, Aral and Taylor randomly applied a single positive or negative vote to comments on a live site and found the initial nudge produced large, persistent differences in final scores, demonstrating that early signals are amplified rather than corrected. Lerman and Ghosh modelled and measured the spread of news items across social platforms and found propagation is governed by the structure of who sees an item early and how quickly they respond, with visibility rules shaping the eventual distribution. Both results describe a system in which the first small sample is treated as an estimate of general appeal and then determines how many further observations are collected.

Why it stays hidden

The dependence hides because the platform reports outcomes and not counterfactuals. An author sees reach and interprets it as feedback on the work. The alternative version — the same item published three hours later — does not exist, so a scheduling variable is experienced as a judgement about quality.