Informational Cascades & Social Norms

Informational cascades were formalised by economists including Sushil Bikhchandani, David Hirshleifer, and Ivo Welch in the early 1990s, and brought into the design conversation by Thaler and Sunstein's Nudge. The idea: people infer what's correct from what they see others doing, then add their own behaviour to the pile.

What looks normal becomes normal. Thaler and Sunstein's treatment of informational cascades shows that people use observable social behaviour as a signal for what's correct or appropriate — and once a behaviour appears widespread, others follow it. This is a more specific and actionable form of social proof: it's not just about showing that someone uses the product, it's about surfacing the prevailing behaviour in a way that makes it feel like the natural thing to do. "Most customers choose this plan" is more powerful than a review count because it directly signals what the right choice looks like.

This is a sharper, more actionable form of social proof. It's not just that *someone* uses the product — it's that a visible, prevailing behaviour signals what the *right* choice looks like, and people cascade behind it. 'Most customers choose this plan' does more than a raw review count, because it tells the visitor what normal, sensible people do.

The same force can run against you. If the visible behaviour around your product looks hesitant or negative — empty activity, sparse adoption — that, too, cascades. What people can see others doing becomes the norm they follow, so it's worth surfacing the genuine prevailing behaviour you want amplified.

An informational cascade happens when people, uncertain what to do, infer the right course from the behaviour of others — and once enough people visibly act a certain way, others follow, treating the prevailing behaviour as evidence. Each person's choice adds to the signal, so a behaviour that looks widespread becomes self-reinforcing. It's a more specific and actionable cousin of social proof.

The distinction matters for design. General social proof says 'others use this'; an informational cascade says 'this is what the right choice looks like, because it's what most people choose'. That's why a marker like 'most customers choose this plan' is so effective — it doesn't just reassure, it directly signals the normal, sensible decision and invites the visitor to fall in line with it.

The lever, used honestly, is to surface genuine prevailing behaviour rather than invent it. Showing what real customers actually choose turns observable behaviour into a guide; fabricating a false consensus is a dark pattern that fails the moment it's doubted. It connects to social proof, the Octalysis social drive, and confirmation bias.

Kweri can note where a page surfaces prevailing-behaviour signals — 'most popular', adoption indicators, what others are choosing — and where an opportunity to do so is being missed in favour of weaker, generic proof. What it can't verify is whether any such claim is true, since that's a fact about your customer base, not the page. So Kweri may prompt you to surface genuine prevailing behaviour where it would help, while the truthfulness of any 'most customers choose this' claim is yours to guarantee. It's designed to encourage real signals, not manufactured consensus.

Identified by Bikhchandani, Hirshleifer & Welch; applied by Thaler & Sunstein (1992). Catalogued from Nudge: Improving Decisions About Health, Wealth, and Happiness.

Formalised in early-1990s economics and applied to choice design in Nudge; the linked summary is the reference used here.

Read the primary source →