The reason the first review shapes every review after it

The first review says four stars and mentions the packaging. Ninety reviews later, packaging appears in a third of them, and reviewers who liked the product give four because five now looks excessive. The product never changed.

What everyone sees

A rating average is treated as an accumulation of independent judgements, gaining accuracy as the sample grows. Independence fails at the second entry. Each reviewer reads what is there, learns which attributes are worth mentioning and where the scale currently sits, and writes into a conversation rather than into a blank form.

What is actually happening

Moe and Trusov modelled ratings forums over time and found that the ratings already posted influenced subsequent ratings and the volume of new ones, producing dynamics that could not be explained by product quality alone — with the social effects also traceable in sales. Schlosser examined what happens when people know their rating will be read and found that anticipating an audience changed the review itself: raters adjusted towards or deliberately against visible opinions, and those who revised in response to others produced systematically different evaluations from those writing privately. By 2026 the effect compounds through summary widgets and generated review digests, which distil the visible set into a phrase that the next reviewer reads before writing.

Why it stays hidden

The dependence hides inside a number that looks like a sample statistic. An average of two hundred implies two hundred observations, and reporting conventions carry no field for the order they arrived in. Any reviewer asked will confirm they wrote what they thought, and they did.