Email Bounce Rate Benchmarks 2026: What Each Source Reports, and Why They Disagree

A good email bounce rate is under 2%, and under 1% is better. That's the line platforms enforce and the band Klaviyo calls healthy. Published "average" bounce rates are a different thing: they range from 2.33% (GetResponse, 2023 data) to 10.09% (Constant Contact, 2026), because each source measures a different customer base with a different definition of a bounce.

This page compares every bounce benchmark we could trace back to the publisher's own page, with the date each figure refers to. If a number circulates on other blogs but isn't on the publisher's current page, we say so rather than repeating it. To work out your own rate from a send export, use how to calculate email bounce rate. To see where each platform takes action, see bounce rate limits by platform. If you want to know your likely bounce rate before sending, run a sample of the list through the free email list health check.

The published figures, side by side

All pages were read on 8 October 2026.

Source	Figure	Period the data covers	What it measures	Link
GetResponse	2.33% average bounce rate	Messages sent in 2023 (2024 report)	"More than 4.4 billion messages sent by GetResponse customers"	GetResponse benchmarks
Constant Contact	10.09% average bounce rate across users	Not stated (2026 post)	Constant Contact's own user data, cited by a community admin	Constant Contact community
Klaviyo	Healthy under 1%; 1-2% room for improvement; over 2% needs attention	Guidance, article updated 10 March 2026	A target band, not an average	Klaviyo help
Mailchimp	No bounce figure shown	Page last updated December 2023	Says it calculated hard and soft bounce rates by industry, but the table shows only opens, clicks and unsubscribes	Mailchimp benchmarks
Campaign Monitor	No bounce figure	2021 data (2022 report)	Opens, clicks, click-to-open and unsubscribes only	Campaign Monitor benchmarks
MailerLite	No bounce figure	December 2024 to November 2025	3.6 million campaigns; opens, clicks and unsubscribes only	MailerLite benchmarks

Two numbers and a target band is less than most bounce benchmark articles show. That's deliberate. Many articles quote hard and soft bounce averages credited to Mailchimp; the current Mailchimp page doesn't display them. Several also cite a Constant Contact knowledge base article on average industry rates; that URL now redirects to an unrelated help page about setting your business industry. We also left out Benchmark Email's bounce figures by company size, because the page gives no date or period for its data.

Why the numbers disagree

A 4x gap between two large platforms isn't noise. It comes from four choices each publisher makes.

1. What counts as a bounce

Constant Contact's bounce report has seven categories, including "Vacation/Auto-Reply". Its own help page says of those messages: "Even though we include these in your Bounced report, they aren't actually bounces." It also includes "Mailbox Full", "Blocked" and "Undeliverable", which Constant Contact retries for up to three days (source). The community post doesn't say which categories went into the 10.09% figure. If the average is built from the same bounce report customers see, it includes messages that were delivered.

GetResponse doesn't define its 2.33% in the report at all. Without a definition you can't tell whether it is hard bounces only, or everything that failed.

2. What it is divided by

A bounce rate divided by messages delivered is always higher than one divided by messages sent, and the gap grows as bounces grow. None of the sources in the table states its denominator. Our calculation guide shows how much this moves the result.

3. Whose customers

Each benchmark describes one platform's customers. GetResponse's own figures show how wide that spread is: its 2024 report lists Israel at 9.97% and Germany at 3.89%, against the 2.33% average. A platform used by small businesses importing old contact lists will average higher than one used by ecommerce brands mailing recent buyers. Neither tells you what your list should do.

4. When

The figures in the table span 2021 to 2026. Lists decay continuously as people change jobs and abandon addresses, so a benchmark from a 2023 send describes a different set of mailboxes from one in 2026.

How to read a bounce benchmark before you use it

Before you compare your rate with any published figure, answer four questions from the source page:

1. Which bounces are counted? Hard only, hard plus soft, or every entry in a bounce report, including auto-replies.
2. What is the denominator? Messages sent, messages delivered, or contacts.
3. Whose sends? One platform's customers, with its own mix of list sources and industries.
4. When? The year the emails were sent, not the year the article was updated.

If the page answers none of these, which is common, the figure is a talking point rather than a benchmark. Your own platform's definition matters more, because that's the one enforcement is based on. Find it in your platform's help centre before you decide whether your number is good or bad.

What the platforms and mailbox providers say

Averages describe what happened. Thresholds describe what happens to you, and those are lower than most of the averages above.

Sending platforms. Published enforcement lines run from Brevo's 2% hard bounce rate to Postmark's 10%, with HubSpot at 5% in a month and Amazon SES reviewing accounts at 5%. Mailchimp and Mailgun say thresholds exist but don't publish them. The full table, read from each vendor's documentation, is in bounce rate limits by platform. For cold email tools specifically, see the Instantly and lemlist guides.

Gmail. Google's sender guidelines publish a spam rate line, "below 0.3%" with a target below 0.10%, but no bounce percentage. On bounces, the guidance is behavioural: "If messages start bouncing or start being deferred, reduce the sending volume until the SMTP error rate decreases."

Yahoo. Yahoo's best practices say to "Monitor hard and soft bounces" and "Remove invalid recipients from your list promptly", again with no number.

Microsoft. The Outlook.com requirements for senders above 5,000 emails a day include a hygiene recommendation to "Remove invalid addresses regularly" (details).

So the mailbox providers that decide inbox placement publish no bounce number at all. They see bounces as one signal of a sender who doesn't maintain the list, and judge it alongside complaints and engagement.

So what number should you use?

Use the threshold, not the average.

Your bounce rate	What it means
Under 1%	Healthy by Klaviyo's definition and under every published platform line
1% to 2%	Acceptable, but find out which segment the bounces come from
2% to 5%	Above Brevo's suspension line and Klaviyo's "needs attention" band; clean the list before the next send
Over 5%	At or above HubSpot's and Amazon SES's review lines; stop and fix the list

Three practical rules follow from the comparison above:

1. Measure hard bounces separately. A 3% rate that is mostly mailbox-full deferrals is a different problem from 3% non-existent mailboxes. The difference, with SMTP codes, is in hard bounce vs soft bounce.
2. Compare yourself with yourself. Your bounce rate on the same list over time tells you far more than another platform's average. A rising trend means the list is decaying faster than you're cleaning it.
3. Fix the cause before the send. Every benchmark and threshold above measures damage after it's done. Most hard bounces come from addresses that no longer exist, and those can be found before you send. Verifying a list removes them (what bulk verification costs), and if a send has already gone wrong, start with what to do after a high bounce rate.

The honest summary of every published bounce benchmark is that none of them is a safe level for your list. They tell you how a platform's customers did in one period. Under 2%, measured against messages sent and tracked over time, is the number that keeps you clear of every published line.