How to track PDF engagement without overstating the data.
Page reach, CTA clicks and form submissions are useful signals, but they are not eye tracking or proof of attention. This guide defines practical metrics and their limits.
By PDFLP · Last updated
On this page5 sections+
Choose a small metric set
- Unique reading sessions.
- Deepest page reached.
- CTA click rate.
- Form completion rate.
- Lead count and lead source.
- Download count after a successful form submission.
Reading depth is a signal, not attention
A page-reached event shows that the browser displayed a later part of the document. It does not show whether the reader understood or valued it.
Use depth together with CTA and form activity. A high reach without action may mean the document is useful but the CTA is weak, or that the page was left open unattended.
Separate reading from lead identity
Reading events use anonymous session identifiers. Contact details appear only after the visitor submits the form.
Do not join unrelated sessions into a personal profile unless the visitor has been informed and the applicable privacy requirements allow it.
Attribution is incomplete
Search engines and AI assistants do not always send a useful referrer. UTM tags can help with campaigns but cannot recover traffic that arrives without them.
A session should be classified as search, AI, direct, referral, paid or unknown only when the evidence supports that classification.
Run useful experiments
- Change one element at a time.
- Compare a direct download with a gated download.
- Test a shorter form against the current fields.
- Review mobile and desktop separately.
- Use a large enough sample before making a permanent change.
- Validate important findings against sales or CRM outcomes.