Bounce Rate: A Practical Guide After the GA4 Change

Quick answer: In Google Analytics 4, bounce rate is the share of sessions that were not engaged, which makes it 100% minus your engagement rate. It is a different metric from the old Universal Analytics bounce rate, so old benchmarks no longer apply. Judge it per landing page, against what that page is meant to do.

Ask a Singapore business owner what their bounce rate is and you will usually get one of two answers: a number someone quoted in a meeting years ago, or a shrug. Both are understandable. The metric most owners learned about came from Universal Analytics, the old version of Google Analytics, and that version no longer exists. Bounce rate still appears in Google Analytics 4 (GA4), but it is now calculated in a completely different way, and almost every rule of thumb you have heard was written for the old definition.

This piece is written for the owner who has GA4 installed and wants to know whether a bounce rate figure is a problem. It covers what the number means since the GA4 change, why “40-60% is normal” is no longer a useful benchmark, how to add the metric to your reports, which single-page visits are perfectly healthy, why WhatsApp and phone taps must count as engagement, and a short routine for finding the pages that actually need work. If you want the wider picture of how measurement fits into search work, our SEO services overview explains where it sits.

Our conclusion is simple: stop judging the site-wide number. Judge each landing page against its job, and spend your effort on pages where engaged visitors still do not enquire.

What Bounce Rate Means in GA4 (It Is Not the Old Number)

In Universal Analytics, a bounce was a single-page session with no further interaction. Someone landed on a page and left without loading a second page or triggering a tracked interaction. It did not matter whether they stayed for two seconds or read for five minutes. If they saw one page and left, it was a bounce.

GA4 starts from a different idea: the engaged session. According to Google Analytics Help, a session counts as engaged if it meets any one of three conditions:

• It lasted longer than 10 seconds (this timer is adjustable, up to 60 seconds).
• It had a key event, which is an action you have marked as important to your business, such as a form submission.
• It had two or more page or screen views.

Engagement rate is the percentage of sessions that were engaged. Bounce rate in GA4 is the percentage of sessions that were not engaged, so the two always add up to 100%. If your engagement rate on a page is 62%, its bounce rate is 38%. There is no separate calculation hiding behind it.

That one change reverses the meaning of many visits. A visitor who reads your entire service page for three minutes, then leaves, was a bounce in Universal Analytics. In GA4 that same visit is engaged, because it lasted longer than 10 seconds. A visitor who taps your phone number within five seconds of landing was also a bounce in the old system unless someone had set up event tracking. In GA4, if that tap is tracked as a key event, the session is engaged.

In GA4, a bounce now means someone arrived, stayed for 10 seconds or less, viewed only one page and did nothing you have told GA4 is important. That narrower definition mostly describes people who left almost immediately, which is worth investigating.

There is also no old data left to compare against. Google stopped processing data in standard Universal Analytics properties from 1 July 2023 and removed access to the remaining properties and their data from the week of 1 July 2024. Any bounce rate you remember from that era cannot be checked or compared any more.

Why “40-60% Is Normal” No Longer Applies

Search for a good bounce rate and you will still find benchmark tables: 20-40% is excellent, 40-60% is average, above 70% is a problem. Those ranges were built on Universal Analytics data, using a definition of bounce that no longer exists. Applying them to a GA4 number is comparing two different measurements that happen to share a name.

Because GA4 counts any session longer than 10 seconds as engaged, pages where people read carefully usually show a much lower bounce rate than they did before. Pages where people make a fast decision, such as tapping a phone number on a contact page, can look worse if that tap is not tracked. The same website can produce a very different GA4 figure from its old one without anything about the site or its visitors changing.

There is a second problem with benchmarks: the timer is adjustable. A property owner can raise the engaged session timer from 10 seconds to 60 seconds, and from then on more sessions will count as bounces. Two businesses with identical visitor behaviour can report quite different bounce rates purely because of one setting. Unless you know how a benchmark was measured, you cannot compare it with your own.

Most agencies will tell you to “get your bounce rate under 40%” and will report a falling number as a win. That frequently backfires, because the easiest ways to push the number down have nothing to do with customers. Splitting one useful page into three thin ones, firing extra events on scroll and marking them as key events, or adding pop-ups that force a click will all lower bounce rate while making the site worse. A falling bounce rate is not proof of progress, and a rising one is not proof of failure.

In our experience, the most useful comparison is not your site against an industry average. It is this landing page against itself over time, and against other pages on your own site that do the same job. A service page’s bounce rate this quarter versus last quarter, measured with the same timer and the same key events, tells you something. A site-wide number against a blog’s benchmark table tells you almost nothing.

How to Add Bounce Rate to Your GA4 Reports

One reason owners never look at bounce rate in GA4 is that it is not shown in most standard reports by default. Google Analytics Help notes that you need to customise a report to add engagement rate and bounce rate. You need the Editor or Administrator role on the property to do this.

The most useful report for this metric is Landing page, under Reports, then Engagement. Bounce rate is about how a visit started, so looking at it by the page people arrived on makes more sense than looking at it by every page viewed. To add it:

1. Open Reports, then Engagement, then Landing page (or Pages and screens if you prefer).
2. Click Customize report (the pencil icon) in the upper right.
3. Under Report data, click Metrics, then Add metric.
4. Type “Bounce rate”, select it, and do the same for “Engagement rate” and “Key events”.
5. Click Apply, then save the changes to the current report.

You can only add metrics to detail reports, not overview cards. If you only see an option to add cards, you are in an overview and need to open a detail report instead.

While you are in the settings, check the engaged session timer. It sits in your web data stream’s tag settings, under the option to adjust session timeout, where you can set the timer for engaged sessions anywhere from 10 to 60 seconds. For most service businesses, leaving it at 10 seconds is fine. Raising it can make sense for long-form content sites, but whatever you choose, write the setting down and do not change it casually, because every comparison over time depends on it.

When we review a client’s analytics, a quietly changed timer is one of the first things we check before reading any engagement trend. If nobody on your team has time to own these settings, our small business SEO page explains how support is usually scoped for businesses without an in-house marketer.

Single-Page Visits That Bounce for Good Reasons

Even under GA4’s narrower definition, some bounces are healthy. The key question is whether the page did its job in under 10 seconds, and for some pages that is exactly what you want.

Think about how people use business websites in Singapore. Someone at a bus stop searches for a clinic’s opening hours, sees them in the first line of the page, and puts the phone away. Someone checks whether a renovation contractor covers their HDB estate, sees the answer on the service area page, and closes the tab to message their spouse. Someone wants a law firm’s address before a meeting, finds it, and leaves. Each of those visits can be over in a few seconds, and each one was a success.

Pages that commonly bounce for good reasons include:

• Contact and location pages, where the visitor needs an address, a map, or a phone number.
• Opening hours and holiday notices, especially around Chinese New Year and public holidays.
• Short answer pages, such as a frequently asked question about fees, parking or eligibility.
• Menu or price list pages that a returning customer opens just to check one item.

On these pages, a high bounce rate can mean the page is clear, not that it is failing. Making visitors click further to find a simple answer would lower the number and annoy the customer.

The other side of this is just as important. A healthy bounce on a short-answer page does not excuse a high bounce on a page whose job is to persuade. If your main service page, the one that should convince a stranger to call you, has most visitors leaving within 10 seconds, that is not people finding their answer. That is people deciding you are not what they wanted. In our experience, owners who worry about the site-wide number often miss this distinction entirely, because the healthy bounces on contact pages and the unhealthy bounces on service pages get averaged into one figure that means nothing.

Track WhatsApp and Phone Taps So Real Enquiries Count

Here is the setup problem that inflates bounce rate on many Singapore sites. A visitor lands on your service page from Google, reads the headline, and taps the WhatsApp button within five seconds. They open a chat, ask for a quote, and become a customer. In GA4, unless that tap is tracked as a key event, the visit may still count as a bounce: one page, under 10 seconds, nothing important recorded.

The same thing happens with click-to-call links (phone number links that start with tel:). On mobile, many customers here prefer to message or call rather than fill in a form. If GA4 cannot see those taps, your best-performing visits, the ones that produced enquiries fastest, are counted among your failures.

The fix has two parts:

1. Create a dedicated event for each enquiry action, for example whatsapp_click when a link points to wa.me or api.whatsapp.com, and phone_click when a tel: link is tapped. This is usually done in Google Tag Manager (a free Google tool for managing tracking code) or with GA4’s own create event feature.
2. Mark those events as key events under Admin, then Data display, then Events. Because a key event makes a session engaged, those quick enquiry visits stop being counted as bounces from that point onward.

GA4 does not rewrite history, so the change only affects data after you make it. Expect your bounce rate on service and contact pages to drop once tracking is fixed. That drop is a correction, not an improvement, and it is worth noting the date so nobody later mistakes it for the result of new work.

When we audited sites whose owners believed their pages “bounced too much”, untracked WhatsApp and phone taps were a recurring cause. The pages were working; the measurement was not. If your business relies heavily on calls and map searches, our local SEO page explains how those enquiry paths fit alongside website tracking.

Bounce Rate by Page Type: What It Means and What to Check

Because every page has a different job, the same bounce rate can mean opposite things on two pages of one website. This table sets out how to read the number by page type, assuming your enquiry actions are already tracked as key events.

Read the table with one rule in mind: bounce rate is only a symptom. It tells you where to look, never what is wrong. The diagnosis comes from opening the page on a phone, checking which searches bring people there, and asking whether the page delivers what those searchers wanted.

Notice that the last column rarely mentions the bounce rate itself. Almost every useful check is about the visitor’s intent, the page’s clarity and whether enquiry actions are tracked. That is why chasing the number directly produces so little. If your sector has its own patterns, such as clinics, law firms or property, our industry SEO page shows how page priorities differ by sector.

Is Bounce Rate a Google Ranking Factor?

No, not in the sense most people mean. Google representatives have said repeatedly that they do not use Google Analytics data, including bounce rate, to rank websites. Gary Illyes of Google said in 2015 that Google does not use analytics or bounce rate in search ranking, and John Mueller said in 2022 that the idea Google looks at analytics bounce rate when ranking websites is a misconception. Plenty of websites rank well without Google Analytics installed at all, and Google has no reliable way to compare bounce rates between sites that measure them with different settings.

This matters because a common sales pitch runs: “Your bounce rate is hurting your rankings, so we need to fix it.” That argument is not supported by what Google has said. Lowering your GA4 bounce rate will not, on its own, move your rankings.

That does not mean visitor behaviour is irrelevant to search. Google’s guidance on helpful content asks whether a page satisfies the people who arrive on it, and a page that consistently disappoints searchers is unlikely to be the best answer for that query. But the route from a disappointing page to weaker search performance runs through the quality and relevance of the page, not through the number in your analytics account.

The practical takeaway is to treat bounce rate as a business diagnostic, not an SEO score. It helps you find pages that are failing visitors, and fixing those pages is good for enquiries whether or not it changes rankings. Speed and mobile usability, which affect how quickly people leave, are separate matters that Google does discuss under page experience, and they are worth fixing for their own sake. If you suspect slow loading or mobile layout problems, our technical SEO service covers how pages are built, loaded and indexed.

When owners ask the Singapore SEO Agency team whether to “fix bounce rate for SEO”, we redirect the question: which pages bring searchers who then do not enquire, and why? That question leads somewhere useful. The other leads to tricks that lower a number nobody at Google is reading.

Case Study: Giving Each Page a Job at a Kampong Glam Boutique Hotel

Our boutique hotel case study documents a 7-month engagement with a 22-room heritage boutique hotel in Kampong Glam, Singapore. The page does not report bounce rate, and we will not invent one. What it does show clearly is the principle behind this piece: each page was built for a specific job, and its success was judged by that job.

The starting position. At month 0 the hotel had 310 organic sessions a month. Its direct booking rate was 26% of total reservations, with an OTA (online travel agency, such as Booking.com or Expedia) commission share of 74%, each booking carrying a 15-18% commission. It was not appearing in the Google Hotels carousel, had 41 Google Business Profile reviews and a Domain Authority (a third-party score estimating a site’s link strength) of 8. The site had no neighbourhood guide, no best rate guarantee landing page and no direct booking incentive.

What the programme did. The page sets out five phases: a Google Business Profile rebuild (months 1-2), a direct booking landing page (months 1-3), neighbourhood authority content covering Arab Street, Haji Lane, the Sultan Mosque area and heritage trails (months 2-5), LodgingBusiness schema (months 2-4), and competitive direct booking SEO pages (months 3-7).

Each page had a different job. The Best Rate Guarantee page was built for guests who had already found the hotel on an OTA. Its job was to convert, and the page reports that conversion from it to direct booking enquiry ran at 8.3%. The neighbourhood guide pages had a different job: to attract travellers choosing by experience, then pass them to room pages through internal links.

The ending position at month 7. Organic sessions rose from 310 to 1,290 a month (+316%). The direct booking rate went from 26% to 72% (+46pp), and OTA commission share fell from 74% to 28%. The page headlines 187% direct booking growth and a 61% reduction in OTA commission spend. Reviews grew from 41 to 118, the hotel began appearing in the Google Hotels carousel for 6 searches, and Domain Authority went from 8 to 21. These results came from the whole programme across all five phases, not from any single page or metric.

The lesson for bounce rate is about method, not the hotel’s numbers: decide what each page is for before you judge how visitors behave on it. Independent hotels weighing a similar approach can see how the work is scoped on our hotel SEO page.

A Monthly Bounce Rate Diagnostic Routine

Once enquiry actions are tracked as key events, you can find problem pages in about 20 minutes a month. The routine below uses the Landing page report with Sessions, Engagement rate, Bounce rate and Key events added.

Step 1: Confirm the measurement has not changed. Check that the engaged session timer is still at the value you recorded, and tap your own WhatsApp button from your phone to confirm it appears in the Realtime report. A broken event can make bounce rate jump overnight with no change in visitor behaviour.

Step 2: Filter to the pages that matter. Ignore pages with very little traffic, where a handful of visits can swing the rate wildly. Focus on your main service pages, your homepage and any page that ranks for searches that should bring customers.

Step 3: Sort into four groups. For each important landing page, note whether bounce rate is high or low and whether key events are high or low. Low bounce with good key events means the page works. High bounce with good key events usually means fast enquiries, which is fine. High bounce with few key events means visitors leave immediately. Low bounce with few key events is the group most owners miss: people stay, read and still do not enquire.

Step 4: Check mobile and speed first. Open each problem page on your own phone over mobile data, not office Wi-Fi. Slow loading, a pop-up covering the headline, or a WhatsApp button buried several screens down are the likeliest reasons for fast exits. Google’s free PageSpeed Insights tool will show whether real visitors experience slow loading.

Step 5: Check intent. In Search Console (Google’s free tool showing which searches bring visitors), look at the queries landing on the page. If people search for prices and your page talks only about your history, they will leave.

Step 6: Fix engaged pages that do not convert. For pages where visitors stay but do not enquire, add a clear offer, area served, proof such as reviews, and an obvious next step near the top. These pages are your best opportunity, because the attention is already there. Our case studies show how Singapore businesses in different sectors structured pages around enquiries.

Step 7: Write one action and one date. Note the change and when it went live, so next month’s numbers can be read against it. If you want to see how ongoing work like this is usually structured before you ask anyone, our pricing page is public.

Field notes: In our boutique hotel case study, a 22-room heritage boutique hotel in Kampong Glam, the page built to convert OTA-aware guests and the pages built to attract neighbourhood searchers were judged by different measures. The Best Rate Guarantee page converted to direct booking enquiry at 8.3%, while the neighbourhood guides fed room pages through internal links. Over the 7 months, organic sessions went from 310 to 1,290 and the direct booking rate from 26% to 72%. The page reports no bounce rate at all, which is the point: the hotel’s progress was read through bookings, not through one engagement number.

Our Take

Bounce rate in GA4 is a narrower, more honest metric than the one most owners remember. It counts sessions that lasted 10 seconds or less, saw one page and triggered no key event. Because the definition changed, the old “40-60% is normal” benchmarks no longer apply, and comparing your GA4 figure with them, or with another business, tells you very little.

Judge each landing page against its job. Track WhatsApp and phone taps as key events so fast enquiries stop counting as failures. Accept healthy bounces on contact and answer pages, and spend your effort on the pages where engaged visitors still do not enquire, because that is where the attention you already earn is being wasted. If you want to understand how we approach this kind of review, our about page explains how we work with Singapore businesses.

Frequently Asked Questions

What is bounce rate in Google Analytics 4?

In GA4, bounce rate is the percentage of sessions that were not engaged. A session is engaged if it lasted longer than 10 seconds, had a key event, or had two or more page or screen views. So a bounce in GA4 is a visit of 10 seconds or less, on one page, with no key event. Bounce rate is always 100% minus engagement rate.

How is GA4 bounce rate different from Universal Analytics bounce rate?

Universal Analytics counted any single-page session with no further interaction as a bounce, regardless of how long the visitor stayed. GA4 counts a session as a bounce only if it was not engaged. A visitor who reads one page for several minutes was a bounce in Universal Analytics but is engaged in GA4. The two numbers measure different things and should not be compared.

What is a good bounce rate in GA4?

There is no reliable universal figure. Most published benchmarks were built on Universal Analytics data, and GA4 bounce rate also depends on your engaged session timer and which actions you track as key events. Compare each landing page with itself over time and with similar pages on your own site, and judge it against that page’s job rather than an industry average.

Why can’t I see bounce rate in my GA4 reports?

Bounce rate is not included in most standard GA4 reports by default. You need to customise a detail report, such as Landing page or Pages and screens, and add Bounce rate as a metric. This requires the Editor or Administrator role. If you only see an option to add cards, you are in an overview report and need to open a detail report instead.

Can I change what counts as an engaged session?

Yes, partly. You can adjust the engaged session timer in your web data stream’s tag settings, from the default of 10 seconds up to 60 seconds. You cannot change the other two conditions, a key event or two or more page views. Changing the timer changes your bounce rate from that point onward, so record the setting and avoid changing it casually.

Is bounce rate a Google ranking factor?

Google representatives have said repeatedly that they do not use Google Analytics data, including bounce rate, to rank websites. Lowering your GA4 bounce rate will not move rankings on its own. Pages that fail searchers are less likely to be the best answer for a query, but that comes from page quality and relevance, not the analytics figure.

Why is my contact page bounce rate so high?

Often because it is working. Visitors come for an address, map or phone number, get it in a few seconds and leave. If phone taps and WhatsApp clicks are not tracked as key events, those successful visits count as bounces. Track them as key events, then check that your address, hours and number are correct before worrying about the figure.

Do WhatsApp clicks reduce bounce rate in GA4?

Only if they are tracked as key events. A session with a key event counts as engaged, so it is no longer a bounce. Create a dedicated event for clicks on WhatsApp links, mark it as a key event, and quick WhatsApp enquiries will stop counting as bounces from then on. GA4 does not apply the change to past data.

Does a slow website increase bounce rate?

It can. If a page takes too long to show useful content on mobile, many visitors leave within the first few seconds, which GA4 records as a bounce. Check problem pages on your own phone over mobile data and run them through Google’s PageSpeed Insights. Pop-ups covering the headline and buttons buried far down the page cause similar fast exits.

Should I try to lower my bounce rate?

Not as a goal in itself. The number is easy to push down with tricks, such as splitting pages or marking scroll events as key events, that make the site worse. Instead, find important pages where visitors leave immediately or stay but never enquire, and fix the reasons. A lower bounce rate may follow, but more enquiries is the result that matters.