Average engagement time per session
Whether a page holds attention, or is somewhere people land and leave.
What it means
Average engagement time per session is user engagement time divided by sessions, shown in minutes and seconds. Engagement time is the stretch when your page was the thing in front of the visitor — the clock stops when the tab goes to the background or the phone is put down. The denominator is every session that included the page, so visits that arrived and left immediately count in the average at close to zero.
Show the math
Formula and a worked example
User engagement time is the total foreground time recorded for the page across the period. Sessions is every visit that included it, engaged or not.
Worked example. A collection page records 60,000 seconds of engagement time across 1,200 sessions. 60,000 ÷ 1,200 = 50 seconds.
Now split those sessions. 800 of them left in under a second and 400 stayed 150 seconds each — 400 × 150 = 60,000, the same total. The page holds the people who stay for two and a half minutes and loses two thirds of its traffic on arrival, and the 50-second average describes neither group.
It answers the question
Did people spend time on this page, or walk straight through it? A page with plenty of sessions and seconds of engagement is a doorway, not a destination — and that's fine for some pages and fatal for others.
Why it matters
It's the one column on the Pages table that measures attention rather than volume. Sessions tell you a page got traffic; this tells you whether the traffic did anything once it arrived, which is what separates a page that needs more visitors from a page that needs rewriting.
It also ranks templates against each other in a way revenue can't. Most page rows carry little or no revenue, so this is often the only signal you have about which of your product or collection templates is doing its job.
What good looks like
There's no target worth chasing, because the right answer depends on what the page is for. A checkout step or a delivery-policy page should be short — someone getting what they need quickly is the point. A long-form product page for a considered purchase should run into minutes, and a few seconds there means the page isn't being read. Compare each page against itself over time and against pages of the same type, never across types. One caution: because non-engaged sessions sit in the denominator at roughly zero, a page can improve for the people who stay and show no movement here at all.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Make the page match the promise that sent people to it | Align the ad, email or search snippet with what the first screen actually shows | Fewer sessions ending on arrival, average up | 1 week | A page written tightly for one promise ranks and matches for fewer queries, so Sessions on that row can fall as the average rises. |
| Fast Fix what loads before the first screen | Compress the hero image and defer render-blocking scripts on the pages taking the most traffic | Average up as more sessions survive the first seconds | 1–2 weeks | Deferred scripts can silence reviews, personalisation and tracking tags. Check what still fires before you trust the improvement. |
| Slow Put the answer people came for above the fold | Move sizing, delivery time and returns up, ahead of the brand story | Longer engagement on product templates | 3–5 weeks | It pushes photography and the buy button lower. On impulse products that can cost you the add to cart you already had. |
| Slow Merge the pages that say the same thing | Collapse near-duplicate collection and landing pages into one stronger page | Engagement concentrates instead of spreading thin | 1 quarter | You lose the long-tail entry points those thin pages provided, and any redirect mistake costs you the traffic outright. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
Traffic and attention are different columns on the same table, and the interesting pages are the ones where they disagree.
A page carrying its weight
More people arriving and each staying longer. Whatever this page does — the layout, the content, the source it draws from — is the thing to reuse on the templates below it.
Small and absorbing
Few visits, and the ones that land stay. The page isn't the problem; its visibility is. This is the cheapest kind of page to grow.
Traffic that doesn't stay
Plenty of visits, none of them lasting. Usually a mismatch between the promise and the page, and occasionally a page that loads so slowly nobody sees it.
Fading on both counts
Fewer visits and shorter ones together points at something breaking rather than drifting — a broken template, a redirect, a change that dropped the page out of navigation.
Two different clocks, and knowing which is running stops a pointless argument. Duration measures a whole visit from its first recorded event to its last, across every page in it. This measures only the foreground time on one page. A visit left open in a background tab inflates one and not the other, which is why the site-wide figure is usually the longer of the two.
Together they separate how many people stayed from how long the stayers stayed. Engaged sessions rising while this average falls means more visits are clearing the bar and stopping just past it — a reach change, not a content change. The average alone reads that as a decline.
Common misreads
Almost nobody experienced 50 seconds. The average is spread across visits that left instantly and visits that stayed for minutes, and on most pages those two groups are what the number is made of.
It isn't. This counts only time your page was in front of the visitor and it's measured per page; the session figure covers the whole visit and keeps counting between recorded events. They move independently and both can be right.
On a checkout step, a size guide or a delivery page, short is the goal — you designed it to be quick. Judge each page against what it's for before you judge it against the row above it.
Also called
Avg. engagement time · engaged time per session · average engaged time
See yoursEngagement time per session for every page and page title, next to the sessions and engaged sessions behind each one.
Open Google Analytics Pages →