Nobody scrolls by accident
Every scroll is a small yes, renewed viewport by viewport. Pages that get read all the way down are the ones that kept asking well.
The scroll is the cheapest gesture on the internet, which fools people into thinking it’s free.
It isn’t. Every scroll is a decision: a small one, made in half a second, mostly below consciousness, but a decision all the same. The reader arrives at the bottom of a viewport and holds a quiet election: keep going, or leave. Winning that election once means nothing. A page has to win it again and again, viewport after viewport, all the way down.
Nobody scrolls by accident, and nobody reads to the bottom out of politeness. If they got there, the page earned it, one yes at a time.
Most pages are designed as if the decision happens once, at the top. Land the hero, nail the headline, and the rest becomes a warehouse where content gets stored in whatever order somebody’s department requested it. You can feel this as a reader: the page opens strong and then flattens into inventory, and somewhere around the third viewport your thumb slows down and you’re gone, not because anything was wrong exactly but because nothing was pulling.
Designing for the next yes
We think of the alternative as momentum design: treating the page as a series of renewed decisions to continue, and designing for each one. Take that seriously and a few things follow.
- End every section on a loose thread. Not a cliffhanger; this is a website, not a soap opera. Just an unanswered implication, a claim still waiting for its evidence, a “how” hanging in the air behind a “what.” Closure is where readers leave; sections that resolve completely are granting permission to stop.
- Put something worth arriving at just below each fold-line, so the gesture keeps paying out.
- Vary the density. A heavy section earns a light one. Text, then air, then image, then text again. Rhythm is a reason to continue all by itself, and monotony is the quiet exit.
- Hold something back. A page that plays its best material in the first viewport is asking people to stay for the credits.
Where momentum gets spent
The inverse list matters just as much: the things that spend momentum.
- Walls of same-density text.
- The premature ask: a “book a call” button arriving before the page has said anything worth calling about, like a handshake that comes with an invoice attached.
- Stock photography interludes that the eye files as filler in a tenth of a second and then learns to expect.
Each one converts a yes into a shrug, and shrugs compound.
Analytics tools call the measurement scroll depth, which makes it sound like a property of the reader. It’s a property of the page.
Depth is granted, not taken.
When most visitors stop at the second viewport, the second viewport is where the page stopped being worth a yes, and no amount of wishing the audience were more patient changes the location of the evidence.
There’s a humility in designing this way that we’ve come to like. You stop assuming attention and start earning it in installments. The reader owes you nothing at the top of the page and somehow slightly less at the bottom, and the layout has to live with that math the whole way down.
Next time a page holds you to the end, scroll back up and count how many times it asked.
Somewhere in that layout, somebody was counting the yeses.
