By Mehdi HassanWhat does scroll depth tell you about a landing page?
Scroll depth tells you what share of visitors reached each part of a landing page. It's a measure of exposure: if only a small share of visitors reach your pricing section, then most people never saw your pricing, however good it is. That makes it useful for deciding what to move up, cut or rewrite.
What scroll depth can't tell you is why visitors stopped. Someone who scrolled a quarter of the way and booked a demo looks the same on a scroll map as someone who scrolled a quarter of the way and left. That ambiguity is the whole topic of this article. For how scroll maps fit with click maps and recordings, see our hub guide to heatmap analysis.
Is low scroll depth good or bad?
Low scroll depth is good when the visitors who stop early are converting, and bad when they're leaving. The number alone can't tell you which, so you need a second signal: conversion, engagement or behavior in recordings.
Scroll pattern | Conversion | What it usually means | What to do |
|---|---|---|---|
Low scroll | High | The hero answers the question and the offer is clear | Leave it; make sure lower sections still support undecided readers |
Low scroll | Low | The hero doesn't make visitors want more, or doesn't match the ad | Rewrite the headline and subheadline; check message match |
Deep scroll | High | Visitors need detail before deciding, and the page provides it | Keep the length; tighten sections that slow people down |
Deep scroll | Low | Visitors are searching for something the page doesn't answer | Watch recordings; look for scrolling up and down |
Sharp cliff mid-page | Any | A section looks like an ending, or a block is dense or off-topic | Fix or move the section just above the cliff |
The table is a starting point for diagnosis, not a rule. Confirm each reading by watching a few recordings of the segment in question before you change anything.
How do you tell whether low scroll is success or failure?
You tell them apart by comparing the scroll depth of visitors who converted with those who didn't. If converters stop early too, the page front-loads the answer well. If non-converters stop early and converters scroll deeper, the hero isn't persuading enough people to keep reading.
Segment by outcome
In GA4, build an exploration that compares scroll events for sessions with and without your key event. Some heatmap tools let you filter scroll maps by visitors who clicked the main call to action.
Segment by device
Run the same comparison for mobile and desktop separately. A hero that fits one laptop screen can take three phone screens.
Segment by traffic source
Paid visitors who clicked a specific ad behave differently from organic readers. Low scroll from paid traffic often means a mismatch between the ad and the page.
Watch the stops
Filter recordings to visitors who left near the cliff. Look for rereading, hovering and quick exits.
Write the diagnosis
One sentence per segment: who stops, where, and what they seem to be missing.
How do you track scroll depth in GA4?
GA4 tracks scroll depth through enhanced measurement, but only one threshold: a scroll event fires when a visitor reaches 90% of the page. For landing page work you'll want more detail, which you add with Google Tag Manager's scroll depth trigger.
- In GA4, check that enhanced measurement is on for your web stream. That gives you the 90% scroll event.
- In Google Tag Manager, create a scroll depth trigger with vertical thresholds at 25, 50, 75 and 90%.
- Create a GA4 event tag that fires on that trigger and sends the threshold as a parameter, for example percent_scrolled.
- Turn off GA4's built-in scroll event if you send your own, so you don't count 90% twice.
- Register the parameter as a custom dimension in GA4 so you can report on it.
- Test in GTM preview mode and GA4 DebugView before relying on the data.
Google's own documentation on enhanced measurement events at support.google.com lists exactly what fires by default. If events aren't appearing, our guide to GA4 not tracking conversions covers the usual causes, and landing page conversion tracking explains how to set up the key event you'll compare against.
Percent thresholds vs section-based scroll tracking
Percent thresholds are easy to set up but drift whenever the page changes length. If you add a section, 50% now lands somewhere else, and week-on-week comparisons break.
Section-based tracking fires an event when a specific section, such as pricing or testimonials, becomes visible. In Tag Manager this uses an element visibility trigger on the section's ID. It takes a little longer to set up, but it answers a clearer question: what share of visitors saw the pricing section?
For pages that change often, we prefer section-based events for the three or four sections that matter most, plus the default 90% event for reference.
How do you read a scroll map?
Read a scroll map by looking for sharp drops in color, then examining what sits right above each one. Gradual fades are normal; cliffs are where something stops people.
The usual causes of a cliff are a section that visually reads as an ending, such as a dark full-width band or a footer-like layout halfway down; a dense block of text or a large table; a section aimed at a different audience; or an early call to action that satisfied the visitor. On B2B pages that serve several audiences, a cliff often sits right after the section for one buyer type, which our guide to a homepage for multiple audiences addresses.
Scroll maps from heatmap tools and GA4 scroll events measure slightly differently, so don't expect identical numbers. Use the map for where and GA4 for trend over time.
What should you change when visitors don't scroll?
When visitors don't scroll and don't convert, change the top of the page first: the headline, subheadline and the first visual. Those decide whether anyone reads on, so lower sections can't help until the hero works.
- Rewrite the hero around the visitor's problem, using their words. If they arrived from an ad, match the ad's promise exactly.
- Show the product early. A real screenshot or short clip gives people a reason to keep going.
- Remove false endings. Break up full-width bands that look like a footer, and avoid large gaps that read as the bottom.
- Move proof up. If testimonials or customer logos sit below the cliff, most people never see them.
- Check speed. Slow-loading heroes lose visitors before they can scroll at all.
If you have the traffic, test the new hero against the old one; our guide to A/B testing landing pages shows how. If not, ship it and compare scroll and conversion over the following weeks.
How does scroll depth relate to bounce and engagement?
Scroll depth, engagement rate and bounce rate are three views of the same question: did the visitor engage with the page? GA4 counts a session as engaged if it lasts longer than 10 seconds (the default, which you can adjust), includes a key event, or has two or more page views. Bounce rate is the share of sessions that weren't engaged.
Scroll adds detail neither metric has: where on the page engagement stopped. A visitor can be engaged by GA4's definition and still never reach your pricing. Our article on bounce rate vs engagement rate explains what GA4 changed, and funnel analysis shows how scroll fits into the wider drop-off picture from visit to signup.
