Events Per Session
How much a visitor actually does on a page, per visit.
What it means
Events Per Session is the average number of events Google Analytics recorded per session on a page. An event is any tracked interaction — a page view, an add to cart, a checkout start, a purchase — so the figure counts activity, not visitors and not pages. It's a plain number, typically between 1 and 10, and it weighs nothing: a page view and a purchase each count as one.
It sits in the Full Page Urls table, which is currently switched off in the app, so you may not see the column on your Pages report yet.
Show the math
Formula and a worked example
Event count is every tracked interaction recorded on that page in the period, added together.
Sessions is the number of visits that included that page — not only the ones that started on it.
Worked example. A collection page takes 8,000 sessions and records 30,400 events: 30,400 ÷ 8,000 = 3.8. A blog post takes the same 8,000 sessions but records 12,000 events: 1.5.
Same traffic, very different behaviour. On the collection page people are filtering, opening products and adding to cart. On the blog post most visits fire a page view and little else.
It answers the question
Do people use this page, or only look at it? Two pages with identical traffic and identical time on page can sit far apart here, and the gap is the difference between reading and doing.
Why it matters
Traffic reports tell you a page is popular; this tells you whether anything happens on it. A page pulling thousands of sessions at barely above 1.0 is being seen and not used — the visitor arrives, the page view fires, and nothing follows.
It's also a fast integrity check. A page sitting flat at 1.0 while similar pages run at 3 or 4 is usually a tracking gap rather than a customer problem, and every conclusion you draw from that page's other numbers is built on missing data.
What good looks like
There's no published band, and there shouldn't be — the right figure depends entirely on what the page is for. Judge each page against pages of the same type: product pages against product pages, policy pages against policy pages. Comparing a checkout step to a blog post tells you nothing.
Within a type, the useful reading is direction. A page that drops after a redesign lost interaction it used to get; a page that rises after one gained it. Treat your site's own median for each page type as the line, and work on the rows sitting well below it.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Check the tracking before the page | Look for pages stuck near 1.0 and confirm events are firing there at all | Either a real fix, or a page you can stop worrying about | 1 week | Fixing tracking changes the number without changing the shop, and it breaks every comparison that crosses the fix date. |
| Fast Put the next step in view | Move the add to cart, the size picker or the filters above the fold on your weakest high-traffic pages | More interactions from the same visits | 1–2 weeks | More clicking is not more buying. You can lift this figure with an interface that makes people work harder for the same outcome. |
| Slow Match the ad to the page | Send each campaign to a page that shows what the ad promised, instead of a general landing page | Visitors who arrive ready to act | 4–6 weeks | Narrower targeting and more landing pages to maintain, and traffic volume usually falls before quality shows up. |
| Slow Take steps out of checkout | Remove fields, screens and optional choices from the pages between cart and payment | This figure goes down on those pages | 1 quarter | It reads as a decline. On a checkout page, fewer interactions per session is the win — which is exactly why the number needs a page type attached before anyone acts on it. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
Activity and time are the two halves of engagement, and they sit in adjacent columns on the same table. Either one alone can be read backwards.
Working the page
Long visits full of interaction. Usually a page doing its job — though a difficult checkout looks identical, so confirm the outcome.
Fast and decisive
Plenty of action in less time. The best shape for a product or collection page: people find what they want and act.
Reading, not doing
Visitors stay a while and interact with almost nothing. Common on content pages that hold attention but never point anywhere.
In and straight out
Short visits with no interaction. Either the page is wrong for the traffic arriving, or nothing on it is being recorded.
Volume beside intensity ranks your work list. A page at 1.2 with 200 sessions is a curiosity; the same figure on your busiest landing page is the most valuable fix on the site. Neither column ranks anything on its own.
One counts how many visits cleared a low bar, the other how much happened in an average visit. Plenty of engaged sessions at a low events figure means people stay but do very little — engagement that never turns into intent.
Common misreads
On a checkout page it usually isn't. Extra interactions there mean extra steps, corrections and retries. Read it against what the page is meant to achieve, not as a score.
A page view is one kind of event among many. Adds to cart, checkout starts and purchases all count too, so this figure runs well above pages per visit on any page where something happens.
1.0 is the floor — one page view and nothing else recorded. On a page where you'd expect more, suspect the tracking before you blame the design.
Also called
Average events per session · events per visit · interactions per session
See yoursYour per-page activity for the period, beside sessions and time on page.
Open Google Analytics →