How PDF analytics work

Once a file leaves your outbox it is silent. Everything sold as document analytics is really analytics about a link.

A PDF cannot tell you anything. What people call document analytics is information about the address the document was opened through, and knowing that changes what you should expect from it.

An open record beside a shared document, showing dates and a count.
An open record beside a shared document, showing dates and a count.

This guide covers what is reliable, what is noise, and how to use it without damaging a relationship.

The file is silent

An attachment, once sent, reports nothing. Not whether it arrived, not whether it was opened, not whether it was forwarded.

Attempts to make files report back have historically been blocked by readers, because a document that phones home is a security problem. Anything that still works is unreliable and depends on the reader's software and settings.

So the practical position is: to know anything, the document has to be opened through an address you control.

What is reliable

It was opened. Yes.

When. Yes.

How many times. Yes, and this is often the more interesting number. A document opened five times over two days is usually being shown to other people.

From how many places. Roughly, which tells you it has been forwarded.

That set is enough to change your behaviour, which is the only reason to measure anything.

Signal Reliability What it is good for
Opened High Timing a follow-up
When High Understanding the process
Times opened High Spotting internal circulation
Pages viewed Low A rough hint
Time on page Low Almost nothing

Page-level data is noisier than it looks

Products present per-page reading time as though it reveals interest. Be careful with it.

A document left open in a background tab accumulates time on whatever page it was showing. A reader who scrolled quickly to page twelve and stopped looks the same as one who read every page.

It is a hint. If eight of ten readers stop at the same page, something about that page is worth looking at. Anything finer than that is over-reading.

And acting on it visibly is worse than useless. Nobody wants to hear that you noticed they spent four minutes on the pricing page.

Per-recipient addresses

If you need to know which person opened it, give each person their own address.

That is honest: each recipient genuinely has their own link, and what you learn is about that link. It also survives forwarding sensibly, because you can see that the address you gave one person is being opened from several places.

The alternative, guessing from timing, is unreliable and occasionally embarrassing.

Three addresses for three recipients, each with its own open record.
Three addresses for three recipients, each with its own open record.

Say it

One line, in the message or on the page, for anything going outside your organisation.

It costs nothing, most recipients assume it already, and the alternative is a conversation you do not want to have with a client who worked it out.

Use it for timing

The good use is narrow and genuinely valuable.

Unopened after three days: send a short reminder, possibly in a different channel.

Opened, no reply: the document did its job and something else is in the way. Ask what.

Opened repeatedly: it is circulating internally. Offer to answer questions.

That is the whole toolkit, and it removes most of the guesswork from following up.

If this is near what you are doing, How document tracking works and How link tracking works cover the cases on either side.

Put it at an address

Share through an address, give each recipient their own, trust opens and discount reading depth, say that it is tracked, and use what you learn to decide when to follow up.

Then following up is a decision rather than a guess.

Questions people ask

Can a PDF itself report anything?

Not reliably. Readers block the mechanisms that would allow it, and a file opened offline has no way to tell anyone. Anything you learn comes from the address it was opened through.

What is reliably knowable?

That the address was opened, when, and how many times. That is the useful part and it is enough to change how you follow up.

Is page-level reading data trustworthy?

Treat it as a hint. A document left open in a tab produces the same signal as careful reading, and a reader who skims to page twelve looks identical to one who read every page.

How do I know which person opened it?

Send a different address to each recipient. That is the honest method, and it means the data is about a link you gave one person rather than a guess.

Should I tell people?

For anything going outside your organisation, yes, in one line. Most people assume it, and being found out having not said it costs far more than saying it.

Keep reading