Short answer: a good bounce rate sits between 26% and 45% for most commercial websites, and anything above 60% deserves a look. But that single number hides more than it reveals. A blog post that answers a question in 90 seconds and an checkout page that loses visitors instantly can both show 70%, and only one of them has a problem.
This guide gives you concrete benchmarks by industry, page type, traffic source and device, then shows you how GA4’s engagement rate completely changes how you should read the number. The goal is simple: stop panicking about a metric that, for some pages, is supposed to be high.
Bounce Rate in 30 Seconds: The Formula and What Changed
Bounce rate is the percentage of sessions on your site where the visitor did not interact in a meaningful way before leaving.
Bounce rate = (Non-engaged sessions ÷ Total sessions) × 100
Universal Analytics vs GA4: two different metrics with the same name
This is the single biggest source of confusion in 2026, and most of the benchmark articles you will find still quote numbers from the old model.
| Universal Analytics (retired) | GA4 (current) | |
|---|---|---|
| Definition | Single-page session with no second hit | Session that was not engaged |
| Time spent | Ignored completely | 10+ seconds counts as engaged |
| Typical value | Higher | Usually 10 to 30 points lower |
In GA4, a session counts as engaged if it meets at least one of these three conditions:
- It lasted longer than 10 seconds (this threshold is adjustable from 10 to 60 seconds in your data stream settings)
- It triggered at least one key event (conversion)
- It included 2 or more page views or screen views
Which means: Bounce rate = 100% – Engagement rate. They are the exact same information viewed from opposite ends. GA4 pushed engagement rate to the front because it is a positive metric that behaves more sensibly, and honestly, that is the number you should be optimising. See https://ezoic.com.
Practical consequence: if your bounce rate suddenly dropped from 68% to 41% when you migrated to GA4, nothing improved on your website. You changed the measuring stick.
What Is a Good Bounce Rate? The General Benchmark
Across the whole of a typical commercial website, here is how to grade the number you are looking at in GA4:
| Bounce rate | Engagement rate | Verdict |
|---|---|---|
| Under 20% | Over 80% | Suspiciously good. Check for duplicate tracking tags first |
| 20% to 35% | 65% to 80% | Excellent |
| 36% to 50% | 50% to 64% | Good, typical for a healthy site |
| 51% to 65% | 35% to 49% | Average, worth segmenting |
| 66% to 80% | 20% to 34% | High, normal for blogs and paid landing pages |
| Over 80% | Under 20% | Investigate: speed, intent mismatch or bot traffic |
Most industry sources converge on the same rough line: 40% or lower is good, 55% or higher starts to look high, and 50% is a reasonable site-wide target. Use that as a sanity check, then throw it away and segment, because the site-wide average is a blended number that describes no real page on your site.
Bounce Rate Benchmarks by Industry
These ranges reflect typical GA4 reporting across commercial websites. Treat the midpoint as the number to beat, not as a pass or fail line.
| Industry | Typical bounce rate | Why |
|---|---|---|
| E-commerce and retail | 25% to 45% | Browsing behaviour naturally creates multiple page views |
| B2B and SaaS | 30% to 50% | Long research cycles, users compare pages |
| Lead generation and services | 35% to 55% | Mix of deep browsers and quick contact seekers |
| Finance and insurance | 40% to 60% | High volume of one-off informational queries |
| Travel and hospitality | 35% to 55% | Comparison shopping drives page depth |
| Real estate | 35% to 55% | Listing browsing is inherently multi-page |
| Health and medical | 50% to 65% | Symptom searches are answered and abandoned |
| Blogs and content sites | 60% to 85% | One article answers the question, job done |
| News and media | 55% to 80% | Social referrals, single-article consumption |
| Local businesses and restaurants | 55% to 75% | Visitors want an address, a menu or a phone number |
| Reference, dictionaries, tools | 65% to 90% | Single-use by design |
Bounce Rate Benchmarks by Page Type (The Segment That Actually Matters)
If you only segment one way, do it by page type. This is where the number becomes actionable.
| Page type | Good bounce rate | Alarm threshold |
|---|---|---|
| Homepage | Under 40% | Over 55% |
| Category or collection page | Under 40% | Over 55% |
| Product page | 30% to 50% | Over 65% |
| Cart or checkout step | Under 25% | Over 40% |
| Pricing page | 30% to 50% | Over 60% |
| Service or solution page | 35% to 55% | Over 65% |
| Blog article (informational) | 60% to 85% | Over 90% with low time on page |
| Standalone paid landing page | 60% to 90% | Judge by conversion rate instead |
| Contact page | 40% to 60% | Over 75% with no form submits |
| Thank you or confirmation page | Can exceed 80%, that is fine | Not a concern |
Pages that are supposed to have a high bounce rate
Nobody says this loudly enough. The following pages have done their job precisely when the visitor leaves:
- Blog posts that answer a specific question. Someone searched “what is a good bounce rate”, got the answer in the first paragraph, closed the tab. That is a win, not a failure.
- Store locator and opening hours pages. The visitor found the address and is now driving to you.
- Contact pages where the visitor clicks a phone number. The conversion happened off-site.
- Single-purpose paid landing pages. They have one CTA and no navigation on purpose. Their KPI is conversion rate, not bounce rate.
- Order confirmation and download pages. The funnel is complete.
- Glossary entries, calculators and tools. One use, one answer.
Fix this first: track outbound clicks, phone clicks, email clicks, form submissions, video plays and scroll depth as key events. GA4 does some of this automatically with enhanced measurement, but not phone or email clicks. Once these fire, sessions that looked like bounces get reclassified as engaged, and your data starts telling the truth.
Bounce Rate Benchmarks by Traffic Source
Traffic quality varies enormously by channel. Comparing a display campaign to an email newsletter on the same scale is meaningless.
| Traffic source | Typical bounce rate | Notes |
|---|---|---|
| 30% to 50% | Warmest audience, should be your best channel | |
| Referral | 35% to 55% | Pre-qualified by the linking site |
| Direct | 40% to 60% | Contains returning customers and untagged traffic |
| Organic search | 40% to 60% | Depends heavily on query intent |
| Paid search | 45% to 65% | Landing page relevance is the main lever |
| Organic social | 55% to 80% | Low-intent scrolling, in-app browsers |
| Paid social | 60% to 85% | Interruption-based, expect high bounce |
| Display and programmatic | 65% to 90% | Accidental clicks inflate this badly |
| AI assistants and chatbots | 35% to 60% | Growing channel in 2026, intent is usually specific |
And by device
- Desktop: the lowest bounce rate, usually 10 to 15 points below your site average
- Mobile: typically 10 to 20 points higher than desktop
- Tablet: usually sits between the two
If your mobile bounce rate is more than 20 points above desktop on the same pages, you have a mobile experience problem, not a content problem. Check load speed, tap target size, intrusive interstitials and font sizing.
How to Read Your Bounce Rate Without Panicking: A 4-Step Diagnostic
- Never judge bounce rate alone. Pair it with average engagement time and key events. High bounce + 3 minutes on page = satisfied reader. High bounce + 8 seconds on page = real problem.
- Segment before you act. Look at bounce rate by landing page, then by device, then by channel. Your worst 10 landing pages by entrances will contain 80% of the opportunity.
- Compare against yourself, not the internet. Your own 12-month trend for a given page type is far more useful than any industry average, including the ones in this article.
- Ask if the bounce cost you money. If a page bounces at 78% but generates 40 qualified leads a month, leave it alone and go fix something that is actually broken.
Red flags: when a high bounce rate is genuinely bad
- Bounce rate above 70% on a homepage, category page or checkout step
- Bounce rate rising sharply after a redesign, migration or template change
- High bounce combined with engagement time under 15 seconds
- A single traffic source bouncing 25+ points above all the others
- High bounce on pages that rank for commercial or transactional keywords
The Real Causes of a Bad Bounce Rate (Ranked by Impact)
- Slow loading. Every extra second before the Largest Contentful Paint costs you visitors. Target LCP under 2.5 seconds and INP under 200 ms.
- Intent mismatch. The page does not deliver what the title, ad or snippet promised. This is the number one killer on paid traffic.
- Aggressive pop-ups. An overlay within the first few seconds on mobile is the fastest way to send someone back to the results page.
- Poor readability. Walls of text, small fonts, low contrast, no subheadings.
- No obvious next step. If the page ends without a relevant link or CTA, the only remaining action is to leave.
- Broken or ugly mobile layout. Horizontal scrolling, overlapping elements, unclickable buttons.
- Bot and spam traffic. Enable bot filtering and check for referral spam patterns.
Quick wins that move the number within weeks
- Compress and lazy-load images below the fold
- Delay pop-ups until 30 seconds or 50% scroll depth, and never on mobile entry
- Add a short answer or summary in the first 100 words of every article
- Insert 3 to 5 contextual internal links per page, placed where the reader is likely to want them
- Add a table of contents on long pages
- Match ad copy, title tag and page headline word for word
- Track phone, email and outbound clicks as key events so real engagement gets counted
Measurement Bugs That Fake Your Bounce Rate
Before you celebrate or panic, rule out these three classics:
- A bounce rate near 0% almost always means the GA4 tag is firing twice on each page load, or a plugin is duplicating the tracking code.
- A bounce rate of exactly 100% on a page usually means very low traffic (one or two sessions), a tagging error, or a page reached only by non-engaging bot hits.
- Sudden drops of 20+ points often trace back to a change in the engaged session time threshold or a newly added auto-firing event.
FAQ: Good Bounce Rate Questions, Answered
Is a 20% bounce rate good?
On paper yes, it is excellent. In practice a site-wide 20% is rare enough that you should verify your tracking first. Duplicate GA4 tags, a double-firing page_view event, or auto-firing custom events are the usual explanation. If tracking is clean and you run an e-commerce or app-like site with lots of internal navigation, 20% is a legitimately strong result.
What is an ideal bounce rate?
For a typical commercial website, 26% to 40% is the ideal band, and anything under 50% site-wide is healthy. For a content-driven blog, an ideal bounce rate is more like 60% to 70% because a single article can fully satisfy the visitor. There is no universal ideal, only an ideal per page type and per channel.
Is a 90% bounce rate bad?
It depends entirely on the page. On a checkout page, a product page or a homepage, 90% is a serious problem that needs immediate investigation. On a glossary entry, a thank you page, a store locator or a blog post reached from social media, 90% can be perfectly normal. Check the average engagement time before deciding: if visitors stay over two minutes, they are reading, not fleeing.
What does a 100% bounce rate mean?
It means every recorded session on that page ended without an engaged interaction: no second page view, no key event and less than 10 seconds spent. Common causes are extremely low traffic volume (a couple of sessions is enough to produce 100%), a missing or misfiring tracking tag, a redirect that breaks the session, or bot traffic. On a genuinely popular page, 100% is nearly always a tracking issue rather than a behaviour issue. Much the same conclusion turns up on optimizely.com.
What is a good bounce rate for a blog?
60% to 85% is normal and healthy. Judge blog performance by average engagement time, scroll depth, newsletter signups and assisted conversions instead. A blog post bouncing at 75% with a 4-minute engagement time is doing exactly what it was written to do.
What is a good bounce rate for a landing page?
Standalone paid landing pages routinely bounce between 60% and 90% because they intentionally strip out navigation. The metric that matters is conversion rate. A landing page that bounces at 80% and converts at 12% beats one that bounces at 45% and converts at 2%.
What is a good bounce rate for an e-commerce site?
Aim for 25% to 45% overall, under 40% on category pages and under 25% on cart and checkout steps. Product pages above 65% usually point to pricing, image quality, stock availability or shipping cost problems.
Does bounce rate affect SEO rankings?
Google has repeatedly said it does not use Google Analytics bounce rate as a ranking factor. However, the things that cause a high bounce rate (slow pages, thin content, intent mismatch, intrusive interstitials) do affect rankings. Fix the causes, not the metric.
Should I use bounce rate or engagement rate in GA4?
Use engagement rate for reporting and goal setting, since it is the default GA4 metric and frames the number positively. Use bounce rate when you need to communicate with stakeholders who still think in the old model, or when comparing against published industry benchmarks. Just remember that they are two sides of the same coin.
The Takeaway
A good bounce rate is 40% or lower for most commercial sites, 50% or lower site-wide is a sensible target, and 55% or higher is worth investigating. But the number only means something once you have segmented by page type, traffic source and device, and once you have paired it with engagement time and conversions. plausible.io published something useful on the subject.
Stop treating bounce rate as a grade for your website. Treat it as a question: did this specific page, for this specific audience, do the job it was built to do? Sometimes the answer is yes even at 85%.