How to See How Long Someone Spent Reading Your Document

Where the number comes from

There are only two ways a hosted document can end up with a duration attached, and knowing which one produced your number tells you more than the number does.

The closing report. When a reader closes the tab, the page gets a brief moment to send a final message saying how long it was open. This is the clean case: one measurement, taken by the thing that was actually there. It is also the case you cannot rely on. The browser decides whether that last message goes out, and it frequently does not — a tab killed by the operating system, a phone that swaps to another app, an in-app browser inside a messaging client that tears the page down without ceremony. None of those are edge cases; they are ordinary mobile reading.

The reconstruction. When no closing report arrives, the duration has to be rebuilt from what was recorded while the document was still open: each page that rendered, and how long it held the screen. Sum those, and you have a session length. This is how most durations in this category exist, ours included — and it is the half of the mechanism a reading-time figure never shows you.

The reconstruction needs bounds, because a page left open overnight would otherwise report a nine-hour read. Ours are deliberately blunt and worth stating: a single page counts for at most 10 minutes, and a whole session for at most 4 hours. Both clamps are rare in practice — 171 of the 22,949 distinct page views on record ever ran past the per-page ceiling, and none of the 3,147 sessions with pages reached the session cap. A tool that reports an eleven-hour read is not measuring more carefully than one that reports ten minutes; it is declining to bound an obvious artefact.

And there is a third outcome, which is the honest one: sometimes there is no duration at all. If nothing rendered, there is nothing to reconstruct from, and the right answer is to record no duration rather than to record zero seconds — a zero would sit in the average and drag it down as though somebody had read for no time. Sharing a document as a tracked link is what makes any of these three outcomes visible to you, and it is free.

How often it is reconstructed

We can put a number on that, because it is recorded per visit on our own production analytics. The result is the reason this page exists.

Two caveats belong next to that, because both cut against reading it as more dramatic than it is. The split counts every recorded session, automated ones included — a scanner that follows a link and renders nothing lands in the 12.6% by design, since filtering it out would hide the exact group this page is about. And the shares describe our corpus and our tracking code; a tool with a different beacon strategy would land somewhere else on the same axis.

The one place this set touches published ground, it agrees with it. Restricted to the older window that whether PDF view counts are real reports on, the no-duration group returns exactly the 436 sessions and 14.4% that page publishes — the same number, not a similar one. That reproduction was run before any of this was written, which is the only reason a fresh cut is worth publishing at all.

What each tool's clock measures

The useful comparison is not accuracy. Every mechanism below is correct about the thing it watches; they differ in what they are watching, and several are not watching reading at all. Google is precise about its own event in a way that is worth quoting, because the precision disappears the moment the number is put in a report:

“when a user clicks a link leading to a file (with a common file extension)”— Google, Analytics Help: enhanced measurement events (file_download)

A click. There is no clock in that sentence at all, which is the point: a great many “engagement” figures rest on an event that has no duration by construction.

Two rows deserve their due rather than a dismissal. Web analytics wins whenever the document is published openly on your own website: it will tell you which page drove the click, and no document tracker can, because the tracker only exists from the link onwards. And DRM wins outright on printing. A licensed viewer installed on the reader's machine can report that a document was printed; a hosted page cannot, because printing happens in software nobody else is in. If paper leaving the building is your question, that is the category to buy — at the price of your reader installing a viewer before they can read anything.

Reading time vs time open

These get used interchangeably and they are not the same measurement. Time open is wall-clock: the document was on screen from here to there, background tabs and interruptions included. Reading time is the sum of the intervals a page was actually the thing in front of the reader. The first is easier to produce and always larger. The second is the one worth acting on, and it is what per-page measurement gives you: five minutes spread across nine pages is a different document from five minutes stuck on page two.

This is also where a single session figure misleads by being technically true. A visit reported as eight minutes could be eight minutes of steady reading, or ninety seconds of reading with the tab abandoned open behind an email client. Only the page-level shape tells you which — the mechanics of that are on how page-by-page PDF analytics actually work. It is also why every figure we publish is a median rather than an average: a handful of very long sessions drags a mean somewhere useless, and those sessions are exactly the ones most likely to be a forgotten tab.

How to see it for your own document

Upload the file, share the link instead of the attachment, and the durations arrive on their own — per page, per visit, with the pages in the order they were read. There is no setup step, no tag to install and no code on anybody's site, because the page serving the document is the thing doing the measuring. You can see the shape of it on the live dashboard running on sample data, without an account.

Be exact about where this stops, because the limit is permanent and it is the same for every tool in the table above. Measurement ends at the download. Once a copy is on somebody else's disk it can be read ten times or never, and nothing here can tell you which — a product claiming otherwise is describing DRM or overstating. A hosted link does not raise that ceiling. It fills in everything underneath it.

Free is not a trial, and per-page reading time is not the part we hold back: page-by-page reading data is on the free plan along with 50 files, 50 live links, the email gate, password, link expiry, download control and twelve months of history. Pro, at $9 a month, is what you buy when the timing of the answer starts to matter more than the answer: an alert the moment a document opens rather than in tomorrow morning's digest, more than 50 files or 50 live links running at once, and twenty-four months of history instead of twelve. The trigger is how much you are sending and how fast you need to react — never the reading data itself, which stays on free.

Keep reading: how page-by-page PDF analytics work, step by step, and how free PDF tracking works, with the limits stated.