Scroll depth tracking: what a presell threshold can tell you

A scroll threshold records a measured position under a particular setup. It does not establish that someone read the argument, saw a specific CTA or completed an order.

Illustration: Two researchers unroll a long paper product page across a studio workbench and mark reading positions with colored clips.
On this page 6 sections
  1. Which event are you actually reading?
  2. What does the threshold mean on this page version?
  3. Did the visitor scroll continuously or use an anchor?
  4. Is your question really about a CTA?
  5. How should a small team choose the next test?
  6. What belongs in the review card?

A reached position is an observation, not a reading score. Before using scroll depth to rewrite a presell page, identify the event, threshold, page version and reporting window. Then state the question it can help investigate. A record can point you toward an anchor or a layout difference without telling you whether the reader believed the argument.

The examples below use an imaginary ceramic cup campaign. They are working notes, not customer results or measured conversion lifts. This guide does not install tracking or connect an account.

Which event are you actually reading?

Check the definition before naming the result. Google Analytics documents its enhanced-measurement scroll event around the first visible 90% vertical depth.1 That is one defined condition, not an automatic report at every quarter of the article. If your report contains several thresholds, identify the custom setup that produced them instead of attributing them to the same built-in event.

Keep the event name, units and firing conditions with the report. Google Tag Manager supports percentage and pixel thresholds and documents once-per-threshold behavior for its trigger.2 These settings describe a measurement choice; they are not evidence of a universal good scroll rate. A missing event can mean the condition was not recorded. It does not by itself explain why.

Illustration: Two researchers unroll a long paper product page across a studio workbench and mark reading positions with colored clips.
The fictional drawn review keeps the page and next step visible; it is not an analytics screenshot.

What does the threshold mean on this page version?

Map the measured position to the actual page. A long explanation and a short product introduction can put entirely different content at the same percentage. Record the version and inspect the relevant viewport before deciding that an argument was reached. The long-form versus short-form guide addresses the editorial choice; depth tracking helps describe the observed path through a particular version.

For a fictional comparison, version A places a materials explanation before the product CTA. Version B removes that section. A matching threshold label would now represent a different point in the story. Do not call the two records equal exposure simply because the percentage is identical. Keep the deleted explanation and changed layout in the comparison notes.

RecordUseful questionUnsupported conclusion
Threshold reachedWhich position was recorded?The argument was understood
Page version changedWhat content moved?Percentages mean the same exposure
Product clickWhich destination was used?An order was completed

Did the visitor scroll continuously or use an anchor?

A jump link can move the viewport to a later section. Test the published presell jump links and record their actual landing points. A later position reached through navigation should not automatically be described as a continuous read of the sections above it. Browser behavior and your event setup need inspection before you explain the report.

Also inspect short layouts. Tag Manager’s documentation explicitly notes that a threshold visible at page load can fire even without physical scrolling.2 Treat that as a reason to check the layout rather than as proof that every recorded event represents an active gesture. Keep desktop and mobile observations identified instead of assuming their visible content is identical.

Interactive

Choose the next depth check

Choose the record you actually have. This review does not measure a visitor or predict a sale.

  1. Which limitation matters for this review?

Without JavaScript, read all four checks below.

Is your question really about a CTA?

If the business question is whether a particular CTA had an opportunity to be seen, define that question directly. A whole-page depth marker may sit before or after the element and can change meaning with viewport size. MDN describes Intersection Observer as a way to observe an element’s intersection with a viewport or ancestor.3 That is a different technical observation from understanding the CTA’s message.

The CTA placement guide helps connect the request to the argument around it. Keep any visibility condition, click and destination separate. An element entering a region does not establish attention; a click does not establish a completed purchase. Do not relabel these records into a single engagement score that conceals their definitions.

How should a small team choose the next test?

Write one question supported by the current evidence. For example: “Does the anchor land after the paragraph the CTA depends on?” That can be checked on the published page before committing to a rewrite. Another question might ask which content moved when a section was shortened. Neither requires inventing a benchmark from somebody else’s funnel.

If you change the anchor, inspect the landing position and reading context again. If you change the explanation, preserve the old version so the editorial difference remains reviewable. If several things change together, name them; do not claim that a depth change isolated one cause. The presell testing guide covers the broader decision framework.

Illustration: Two researchers unroll a long paper product page across a studio workbench and mark reading positions with colored clips.
A concrete page check can resolve a navigation question before the team makes a wider story change.

What belongs in the review card?

Record the exact URL and version, event definition, date window and the comparison conditions you actually know. Write missing evidence as unknown. Keep an anchor observation separate from an aggregate report and a product click separate from an order. This makes the next discussion about a defined question instead of an attractive percentage.

The original-English Chrome for Developers video below shows Device Mode in its 2022 interface. It provides context for inspecting different simulated viewports, not a scroll-tracking setup or an analytics report for this page. Do not treat a simulated view as a distribution of real visitor experiences.

Analytics Mania demonstrates scroll-depth reports in Google Analytics 4. Use the report to locate drop-off, then check the CTA and product handoff separately. Video: How to see Scroll Depth in Google Analytics 4?, by Analytics Mania - Google Analytics & Tag Manager, on YouTube. It loads from YouTube only when you press play.

Interactive

Write a depth-review card

Use fictional labels. This local note builder installs nothing and saves no report.

  1. Review scopePage: Fictional cup presell, version A. Evidence: Built-in scroll event; definition checked. Setup: Same page version, viewport group and date window.
  2. Next checkInvestigate: Whether an anchor explains the reached position. Verify handoff: Product-page click recorded separately. Keep unknown: Comprehension, intent and order outcome.

Before depth becomes a page verdict

5 items

You can draft a page in ClickBanger and inspect its published route separately. These local review aids do not configure analytics or change the app’s statistics.

Frequently asked questions

Does a scroll event prove someone read the page?

No. It records the configured position or visibility condition, not comprehension. Keep reading, clicking and ordering as separate questions.

Does GA4 automatically supply every quarter-page threshold?

The enhanced-measurement scroll event described in Google’s documentation uses the first visible 90% vertical depth. Additional threshold series require a separately defined setup.

Can an anchor change the depth record?

Yes, an anchor can move the viewport to a later position. Check that path rather than treating every reached position as continuous reading.

Can I compare percentages after deleting half the article?

Keep both versions labeled and inspect which content each threshold represents. Equal percentages do not establish equal exposure.

Does a short page require physical scrolling to hit a threshold?

Not necessarily. Google’s Tag Manager documentation says a threshold already visible at load can fire without physical scrolling.

Do the local tools install tracking?

No. They organize a review using fictional labels. They do not connect analytics, change settings or collect visitor behavior.

Sources

  1. Enhanced measurement events, Google Analytics Help
  2. Scroll Depth trigger, Google Tag Manager Help
  3. Intersection Observer API, MDN Web Docs

ClickBanger Editorial We build presell pages for e-commerce brands and spend our days reading ads, landers and test results. Questions or a correction? Write to us.

Keep reading

All articles

Testing & growth5 min read

Presell jump links: help readers reach the right section

Presell jump links should take readers to the section named in the label. Use them when a specific answer is hard to reach, then test the target, sticky header and keyboard flow. Keep navigation distinct from a purchase action, and preserve a useful route without JavaScript.

Testing & growth15 min read

Presell page CTA placement: choose the next useful step

Place each presell CTA where the reader has enough proof to take that step. Offer a clear first action, repeat the same product destination after useful evidence, and test a sticky bar separately. Measure completed purchases as well as clicks before keeping a new placement.

From product URL to presell page in minutes

Advertorial, listicle, quiz, comparison and 12 more formats, written from your product page and ready to edit.