93% of Users Never Scroll Past the Third Result Card

93% of users never scroll past the third result card. That number should bother anyone who designs interfaces for a living, because it means most of your careful work on cards four through twenty is decoration. The interesting question isn't whether the stat is exactly right — it's why we stop there, and what that says about how people actually make choices when they're not sure what they want.

Your users aren't lazy, they're satisficing

Herbert Simon coined the term "satisficing" back in the 1950s: when the cost of evaluating more options outweighs the expected gain, people take the first option that clears their internal bar and move on. Scrolling is work. Each additional card demands parsing, comparing, and holding the previous ones in memory. Somewhere around card three, the mental ledger tips and the brain says good enough.

This isn't a failure of attention. It's a rational response to a badly priced menu of options. If your fourth card is only marginally better than your first, you've made scrolling expensive and unrewarding at the same time.

Variable rewards keep people pulling, but only if the pull is cheap

The slot-machine metaphor gets overused, but the underlying mechanism — variable-ratio reinforcement — is well documented and shows up everywhere from email refresh to infinite scroll. Unpredictable payoff produces more persistent behavior than predictable payoff.

Here's the catch for designers: variable rewards only sustain behavior when the cost per attempt is near zero. A pull-to-refresh costs nothing. Scrolling through a dense card grid costs attention, working memory, and time. If your layout makes each card expensive to evaluate, you've built a variable-reward system with a punishing entry fee. People will take the first decent hit and quit.

The third card is where loss aversion kicks in

Kahneman and Tversky's work on loss aversion tells us losses feel roughly twice as heavy as equivalent gains. Once someone has invested attention in cards one through three, they've built a small mental model of what's on offer. Scrolling further means risking that model — you might find nothing better, and now you've spent the effort for a net loss.

The counterintuitive fix isn't to make more cards. It's to make the decision cheaper. Show comparison signals early: price, rating, delivery time, whatever your equivalent is. If card two already tells someone how it differs from card one, they don't need to scroll to feel confident.

A concrete example: Booking.com's card density

Booking.com ran years of A/B tests on their search result cards and landed on a layout that front-loads the decision signals — price, review score, location, cancellation policy — into a compact block. They also surface a "deal" badge on a rotating subset of properties. That rotation is variable-ratio reinforcement in plain sight. But it works because the card itself is scannable in under two seconds. The reward is cheap to discover.

Contrast that with a marketplace I audited last year where each listing card carried eleven lines of metadata. Users scrolled an average of 1.4 cards before clicking. Same inventory, better prices further down, nobody ever saw them.

Design for the third card, then earn the fourth

Stop treating the feed as a catalogue and start treating it as a decision surface. The first three cards should contain everything a reasonable person needs to commit. Anything after that has to justify its own existence — either by offering a genuinely different class of option, or by being so cheap to scan that scrolling feels free.

The forward-looking move is to instrument this. Track scroll depth against click-through, not as a vanity metric but as a diagnostic. If your CTR collapses after card three, you don't have a discovery problem. You have a cost-per-card problem. Fix that, and the scroll will take care of itself.