Time per page tells you where a reader’s browser tab stayed visible and active, so it shows attention, not agreement or understanding. A long time on the pricing page can mean serious comparison, confusion, a call with a colleague or a coffee break. Treat it as a reason to ask a better question in your follow-up, never as a conclusion about what the client thinks.
Used that way, page-level data on proposals and PDFs is genuinely helpful. Used as a mind-reading tool, it leads to awkward follow-ups and wrong forecasts. This article explains how the measurement works, which patterns are worth noticing and where the conclusions stop.
Why an attachment cannot tell you anything
When you attach a PDF to an email, the recipient downloads a copy and opens it on their own device, in whatever program they like. Nothing reports back. Even a tracking image in the email only tells you something about the email, not the attachment.
To see page-level activity, the document has to be shared as a link and viewed in a browser page that can measure what happens. That brings practical benefits besides measurement: you can set the link to expire, limit the number of views and replace an outdated version without chasing old attachments.
How time per page is measured
A browser can report two things reliably: whether the tab is visible to the user, and how much of an element is on screen. These are standardised by the W3C as the Page Visibility and Intersection Observer specifications. A well-designed document viewer combines them with recent activity, such as scrolling or moving the pointer, to decide whether a second counts.
Without these rules the numbers are close to useless. A proposal left open in a background tab over lunch would show forty minutes on page 3. With them, time accrues only when:
- the tab is visible and focused,
- the page is substantially in view (for example, at least half of it),
- the reader was active recently.
Even then, the result is attention in a browser. It does not tell you that the person read the words, understood them or liked them.
Patterns worth noticing, and what they can mean
| What you see | Possible meanings | What it does not tell you |
|---|---|---|
| Long time on the pricing page | Comparing options, checking against a budget, confusion about a line item | Whether the price is acceptable |
| Short session, mostly on page 1 | Quick check that the document arrived, skim for later | Lack of interest |
| Several sessions over a few days | The proposal is being considered, possibly discussed internally | That a decision is near |
| Jump straight to terms or the last page | Looking for conditions, signature block or next steps | Agreement with the terms |
| A likely forward to someone new | More stakeholders, perhaps finance or a manager | Who that person is or what they think |
| No session at all | Not opened yet, opened elsewhere, or the reader declined measurement | That the proposal was ignored |
The useful habit is to look for combinations rather than single events. Several sessions, a forward and time on the pricing page together suggest a proposal that is being taken seriously by more than one person. A single long visit to one page suggests very little on its own.
An example
You send a proposal on Tuesday. On Wednesday the client spends a few minutes on the overview and pricing pages. On Friday there is a second session, flagged as a likely forward, again on pricing and then on the payment terms.
A reasonable reading: the proposal has reached someone else, possibly a budget holder, and pricing and terms are the focus. A reasonable follow-up on Monday: “Would it help if I joined a short call with you and your colleague to go through the options and payment terms?” You ask about the likely topic without revealing what you measured.
An unreasonable reading: “They spent ages on pricing, so they think it is too expensive; let me offer a discount.” Nothing in the data supports that.
Common ways to overread page data
- Equating time with interest. Dense tables take longer than a cover page. A complicated page can attract time because it is unclear.
- Ignoring reading habits. Some people read on paper, some forward and let others read, some decide in a meeting.
- Reading too much into absence. A reader can decline measurement, use another device or read a printed copy.
- Using it as evidence. Viewing data is not proof that a client read, understood or accepted a document.
- Saying it out loud. “I noticed you spent four minutes on page 6” rarely helps a relationship.
Questions to ask before acting on page data
Before you change a price, chase a client or update a forecast because of what a proposal viewer showed, run through a short check:
- Is there more than one signal? Sessions on different days, a forward or a question carry more weight than one visit.
- Could the page simply be hard? If people linger on the same page in many proposals, the page may need rewriting rather than the client needing a call.
- Did the reader agree to measurement? No data can mean a declined consent, not a lack of interest.
- Would I be comfortable if the client saw my reasoning? If the follow-up only makes sense because of tracking details, rephrase it as an open question.
- What did the client actually say? A reply or a question always outranks viewing data.
These questions take a minute and prevent most of the classic mistakes: discounting too early, pushing too hard and reading silence as rejection.
Designing proposals that give clearer signals
You get more useful data from documents built for it:
- Put pricing on its own page, and terms on another, so attention is not blended.
- Keep an executive summary on page 1 for readers who skim.
- Offer options (for example, two or three packages) so that the follow-up question is natural.
- End with one clear next step, such as accepting, asking a question or booking a call.
Proposal and PDF tracking in HawkOpen
HawkOpen, the email, document and proposal tracking app made by TriStiX S.L., includes PDF links and proposal links. They are currently a preview: HawkOpen accepts files only once an external malware-scanning service is switched on, and until then uploads are refused. Email open and click tracking does not depend on it.
When file scanning is on, this is how measurement works:
- Consent on the page comes first. Time on document pages is measured only after the reader agrees on the PDF page. Until they answer, and after a refusal, HawkOpen measures nothing, shows the sender no session and keeps no IP address.
- Strict counting. For each viewing session HawkOpen adds up active time per page. Time counts only while the tab is visible and focused, the page is at least half in view and the reader was active in the last 30 seconds.
- Access control. The recipient confirms access with a one-time code sent by email. Links can expire and the number of views can be limited.
- Proposals. A proposal link is created per recipient, shows time per page, flags a likely forward and offers client actions such as accepting, asking a question or requesting a meeting, each confirmed with a receipt signed by HawkOpen. Link previews by bots are not counted as views.
HawkOpen is in a controlled launch. For the details and limits, see the FAQ on which PDF page was read longest and the overview of proposal tracking. If the next step after a proposal is a signature, our article on simple e-signatures and “I accept” clicks explains when a click is enough.