Bounce Rate
Tells you whether the page your traffic lands on deserves the click.
What it means
Bounce Rate is the share of visits that ended without the visitor doing anything the tracking counts — no second page, no key event, and under 10 seconds on site. It's a percentage of sessions rather than of people, so one visitor arriving twice is counted twice. Lower is better here, so a rise is the bad direction.
Show the math
Formula and a worked example
A visit is one session — one person, one arrival, however many pages follow it.
Engaging is the analytics definition, not yours: the visit lasted more than 10 seconds, or fired a purchase or key event, or reached a second page. Miss all three and it counts as a bounce.
Worked example. A landing page takes 10,000 visits in a month and 4,200 of them leave without engaging. Bounce Rate = 4,200 ÷ 10,000 = 42%. If that traffic came from ads at €0.40 a click, those 4,200 bounced visits cost you €1,680 and returned nothing.
The 10-second rule is why a fast, useful page can still look bad. Someone who finds your returns policy in 8 seconds and leaves happy counts against you.
It answers the question
Did the visitors you paid for stay long enough to see anything? A high rate means you're paying for clicks that never turn into engaged visits.
Why it matters
Every bounced visit was paid for. An ad click, an email send or years of SEO put that person in front of the page, and the page gave nothing back.
It's also the fastest warning you get. Bounce Rate moves within hours of a new campaign, a theme update or a heavier hero image, while sales take days to show the same damage.
What good looks like
These bands assume a mixed-traffic store. A page that answers one question can bounce at 60% and still be doing its job — someone read your delivery times and left satisfied. Compare like pages with like, and treat any sudden jump as a tracking change until you've ruled that out.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Send each ad to the page it promised | Point every campaign at the product or collection in the creative, never the homepage | Bounce Rate down 5–10 points on paid traffic | 1 week | You now maintain a landing page per campaign, and visitors see less of the rest of your range — pages per visit falls along with the bounce. |
| Fast Lighten the first screen | Compress the hero image, delay the pop-up, cut anything that loads before the page is readable | Bounce Rate down 3–6 points | 1–2 weeks | A delayed pop-up collects fewer email sign-ups. That shows up weeks later as a smaller list and less revenue from flows and campaigns. |
| Slow Rewrite the top of your 5 biggest landing pages | One clear promise, the price, and a single call to action visible without scrolling | Bounce Rate down, more product views per visit | 3–5 weeks | Design and copy time, and a sharper page narrows the appeal — the browser who wanted to see the whole range leaves earlier. |
| Slow Buy different traffic | Move budget from broad interest audiences towards search and retargeting | Bounce Rate falls without touching the site | 1 quarter | Cheap reach disappears. Sessions drop and cost per click rises — you improved the ratio by buying fewer, better visits. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
On its own it tells you a page lost people. Next to Sessions, it tells you whether to buy more traffic or fix the page first.
Growing on quality
More people arrived and a bigger share of them stayed. The new traffic matches what the page offers — the one combination that's safe to scale.
Volume without interest
You bought reach, not intent. Normal for the first week of a broad campaign, a problem if it's still true in the second.
Smaller and sharper
Fewer visits, better matched. This is what turning off a cheap source looks like — the ratio improved because the weak traffic left.
Something broke
Fewer people coming and more of them leaving on arrival. This is rarely a marketing problem, so look at the site itself first.
A high CTR (Link) with a high Bounce Rate is the clearest sign your ad and your page disagree: the ad is winning the click on a promise the page doesn't keep. Neither number shows that alone — CTR (Link) looks like a win, and Bounce Rate looks like a site problem.
Low bounce with a weak Purchase To View Rate means the page holds attention and still doesn't sell, which is a copy and pricing problem rather than a traffic one. High bounce with a strong Purchase To View Rate is the reverse: the page converts whoever stays, so the fix is the targeting. Neither number tells you which of the two you're looking at.
Common misreads
A number that good is usually a measurement fault — a duplicated tag or a plugin firing an extra event marks nearly every visit as engaged. Check a page you know is weak; if it also reads 12%, you're looking at the tracking, not the store.
Check what changed upstream first. One new campaign running at 70% can lift the site-wide number while every page stays exactly as it was. The page is guilty only if its own daily figure moved.
It means nothing got recorded. Someone who checks your opening hours and leaves got what they came for, so on pages that answer one question, judge the number against that page's own history rather than a target.
Also called
Engagement rate, inverted · single-page session rate · bounce %
See yoursYour Bounce Rate for the period, with the trend against the period before it and a day-by-day breakdown underneath.
Open Google Analytics Overview →