Benchmark · Email Marketing

Email bounce rate benchmarks

Email bounce rate, current figures for total bounces: 0.79% to 0.84%, from 2 sources. We checked 8 sources and quote figures from 5. Email bounce rate is the share of emails sent that could not be delivered and bounced back to the sender. Here is every published figure we can show, side by side, with what each one counts.

5 sources quoted Grade B Checked Nov 12
The takeaway
Published email bounce rates run from

0.01% to 5.74%

.

Published email bounce rates run from 0.01% (Mailgun's Air Freight & Logistics senders, 2025) to 5.74% (Benchmark Email's undated average for companies with 100 to 250 employees). Dotdigital is the only source we can quote that splits the total: in its 2026 data, 0.79% of emails sent bounced worldwide, 0.74% soft and 0.05% hard. Mailgun and Benchmark Email each publish a single bounce rate without saying what it counts. 20 figures from 5 sources, grouped by what they count (hard bounces, soft bounces is charted too, but not part of the headline range). We never average them.

Current · Total bouncesFull published spread 0.01%–5.74%

Current = dated, current editions; older, undated and top-10% figures are excluded (greyed out in the chart).

Current total bounces figures run from 0.79% to 0.84%; the full published spread is 0.01% to 5.74%.

0.79–0.84%
Current total bounces

Where the 2 current total bounces figures from 2 sources sit. Older, undated and top-10% figures are charted separately.

2 sources · 2026
−0.52 pts
Calculated gap (Americas): Hard − Soft
Same sends: soft, hard and total bounces

Dotdigital publishes all three for each region. Soft bounces make up almost all of the total.

5 sources
Grade B, never averaged

Every figure comes from one named source, shown with its sample, data window and statistic.

20 figures · checked Nov 12
See all 20 figures, grouped by what they countGreen band = current total bounces range

Published email bounce rate figures by way of counting, from 0.01% to 5.74%. Every figure is listed in the source list below.

Total bounces2 figures · 2 sources
0.79–0.84%
Bounce rate, not defined9 figures · 2 sources
0.01–5.74%
Hard bounces3 figures · 3 sources
0.04–0.06%
Soft bounces3 figures · 3 sources
0.56–0.79%

Tap or hover a point to see its source.

Current figure Older, undated or a top-10% tierEvery figure is also in the source list below.

Key findings

Each one cites its source figures.

LEAD FINDING
0.79%

Dotdigital's 2026 global average total bounce rate: 0.74% of emails sent soft-bounced and 0.05% hard-bounced. By region the total runs from 0.75% (APAC) to 0.84% (EMEA).

0.01%–0.22%

Mailgun's 2025 bounce rates for its ten highest-volume industries, from more than 400 billion emails. The page does not say whether soft bounces are included.

Source · Sinch Mailgun
1.05%–5.74%

Benchmark Email's averages by company size, the highest figures we can show. The page carries no date and gives no overall bounce rate.

Source · Benchmark Email

Same data, different answers

How the way of counting changes the number, on one scale.

Same sends: soft, hard and total bounces

Dotdigital publishes all three for each region. Soft bounces make up almost all of the total.

Global
Soft0.74%
Hard0.05%
Total0.79%
Source · Dotdigital
EMEA
Soft0.79%
Hard0.06%
Total0.84%
Source · Dotdigital
APAC
Soft0.71%
Hard0.04%
Total0.75%
Source · Dotdigital
−0.52 pts
Calculated gap: Hard − Soft
Americas
Soft0.56%
Hard0.04%Printed total (0.06%) does not reconcile; not shown
Source · Dotdigital

Source by source

Open any source for its definition, sample and method.

DotdigitalEmail and SMS marketing benchmarks 2026: Americas edition · 2026 edition (public regional post; full Global Benchmark Report is gated)0.56%soft bounces, AmericasGrade B
What counts
Soft, hard and total bounce rates, each divided by emails sent. The printed delivery-rate formula reads 'emails sent ÷ delivered emails', the inverse of a delivery rate.
Sample
Over a billion real marketing campaign sends
Data window
Not stated
Figures
  • 0.56% soft bounces, Americas
  • 0.04% hard bounces, Americas
Statistic
Averaged ('anonymized and averaged'); method of averaging not stated
Checked
Nov 12, 2026 · Read the source
Method
Anonymised Dotdigital customer campaign data, averaged; formulas given for every bounce metric; regional editions for Americas, EMEA and APAC.
Published cuts
region (Americas) · bounce type (soft, hard, total) · email vs SMS
Limitations
  • Data window not stated
  • The printed total bounce rate (0.06%) is lower than the printed soft bounce rate (0.56%), so the total is not shown here
  • Delivery-rate formula printed inverted
  • Vendor-reported; enterprise/retail customer mix
Licence
Website terms (updated 2019-01-16) assert ownership of site content but contain no consent requirement or ban on citing figures. Regional figures quoted with attribution.
DotdigitalEmail and SMS marketing benchmarks 2026: EMEA edition · 2026 edition (companion regional post to the Americas edition)0.84%total bounces, EMEAGrade B
What counts
Same formulas as the Americas edition: soft, hard and total bounces divided by emails sent.
Sample
Over a billion anonymized campaigns (shared Dotdigital dataset)
Data window
Not stated
Figures
  • 0.84% total bounces, EMEA
  • 0.79% soft bounces, EMEA
  • 0.06% hard bounces, EMEA
Statistic
Averaged (method not stated)
Checked
Nov 12, 2026 · Read the source
Method
Regional cut of Dotdigital's 2026 benchmark dataset.
Published cuts
region (EMEA) · bounce type (soft, hard, total)
Limitations
  • Data window not stated
  • Delivery-rate formula printed inverted
Licence
Same Dotdigital website terms (silent on citing figures).
DotdigitalEmail and SMS marketing benchmarks 2026: APAC edition · 2026 edition (companion regional post)0.79%total bouncesGrade B
What counts
Same formulas as the Americas edition, with global averages for soft, hard and total bounces.
Sample
Over a billion campaigns (shared Dotdigital dataset)
Data window
Not stated (year-over-year changes published)
Figures
  • 0.79% total bounces
  • 0.74% soft bounces
  • 0.05% hard bounces
Statistic
Averaged (method not stated)
Checked
Nov 12, 2026 · Read the source
Method
Regional cut of Dotdigital's 2026 benchmark dataset, with global averages for comparison.
Published cuts
region (APAC) · global average · bounce type (soft, hard, total)
Limitations
  • Data window not stated
  • APAC delivery rate given only as 'over 99%'
Licence
Same Dotdigital website terms (silent on citing figures).
Sinch MailgunThe 2026 email impact report: Email benchmarks: How do you measure up? · 2026 edition (2025 sending data)0.22%bounce rate, not defined · Professional ServicesGrade B
What counts
'A bounce happens when an email can't be delivered.' Hard bounces 'mean the address doesn't exist'; soft bounces are temporary. Whether the rate counts both, and its denominator, are not stated. Delivery rate is 'the percentage of emails accepted by receiving mail servers'.
Sample
More than 400 billion emails sent through Sinch Mailgun across its top 10 industries by send volume in 2025
Data window
Calendar year 2025
Figures
  • 0.22% bounce rate, not defined · Professional Services
  • 0.18% bounce rate, not defined · Retail
  • 0.14% bounce rate, not defined · Information Technology
Statistic
Not stated (one rate per industry)
Checked
Nov 12, 2026 · Read the source
Method
'Every number in this table comes from real Sinch Mailgun sending data, not surveys, or estimates.' Delivery rate, bounce rate and unsubscribe counts for the ten highest-volume industries. The rest of the report is a survey of 1,234 senders (late September to early November 2025).
Published cuts
industry (top 10 by send volume)
Limitations
  • No all-industry figure
  • Bounce-rate formula, denominator and hard/soft split not stated
  • Delivery and bounce rates do not add up to 100% (Information Technology: 98.05% and 0.14%)
  • Media's delivery rate is printed as 95.95% in the table and 95.05% in the text
  • Senders on an email API service; the mix of transactional and marketing email is not stated
Licence
Sinch Email Terms of Service (last revised 2025-02-15) apply to Services ordered under a Service Order; they contain no clause on website content, copying or citing, and only the name and logo need prior written consent (s.12). Silent on citing: a few figures, attributed and linked.
Benchmark EmailEmail Marketing Benchmarks By Industry and Region · Undated5.74%bounce rate, not defined · 100-250 employeesGrade B
What counts
'Average Bounce Rate' by employee-size bracket; what counts as a bounce and the base are not stated.
Sample
68,000 users in over 90 industries across 21 regions
Data window
Not stated
Figures
  • 5.74% bounce rate, not defined · 100-250 employees
  • 5.21% bounce rate, not defined · 000 employees
  • 4.65% bounce rate, not defined · 0-25 employees
Statistic
Average
Checked
Nov 12, 2026 · Read the source
Method
Benchmark Email's own user data; bounce and abuse rates are published only in the employee-size table.
Published cuts
company size (employee brackets)
Limitations
  • No publication date or data year
  • No overall bounce rate and no hard/soft split
  • Industry and region tables carry no bounce column
  • Vendor-reported; SMB skew
Licence
Terms of use (updated 2026-07-10) cover the platform; silent on citing website content. Overall figure quoted with attribution.
Show all 6 sourcesShow fewer sources
OmnisendEmail marketing benchmarks: What good looks like in 2026 · 2026 article (2025 data)Context sourceGrade B
What counts
Publishes a 'deliverability rate' for campaigns (98.4%), undefined, and no bounce rate.
Sample
27,000+ brands; more than 20 billion campaign emails and 470 million automated email sends
Data window
Calendar year 2025
Statistic
Not stated
Checked
Nov 12, 2026 · Read the source
Method
Omnisend's own 2025 sending data for a 27,000+ brand set.
Published cuts
campaigns
Limitations
  • No bounce rate published
  • Deliverability rate not defined
  • E-commerce senders only
Licence
Same Omnisend terms as the report (restriction targets 'Omnisend Content' provided as part of the Services). A few figures, attributed and linked.
2 more publishers, linked but not quoted

These publishers' terms do not allow their figures to be reproduced without permission (or we could not confirm that they do), so we link to their reports instead of showing the numbers.

By industry and region

Selected figures, each with its source. Full tables live in each publisher's report.

By industry
Professional ServicesSinch Mailgun · bounce rate, not defined0.22%
RetailSinch Mailgun · bounce rate, not defined0.18%
Information TechnologySinch Mailgun · bounce rate, not defined0.14%
MediaSinch Mailgun · bounce rate, not defined0.08%
Air Freight & LogisticsSinch Mailgun · bounce rate, not defined0.01%
By region
APACDotdigital · total bounces0.75%
APACDotdigital · soft bounces0.71%
APACDotdigital · hard bounces0.04%
EMEADotdigital · total bounces0.84%
EMEADotdigital · soft bounces0.79%
EMEADotdigital · hard bounces0.06%
AmericasDotdigital · soft bounces0.56%
AmericasDotdigital · hard bounces0.04%

Why sources disagree

Each factor moves the number on its own.

99.16%

of Dotdigital's EMEA emails were delivered and 0.84% bounced, which adds up to 100%. Its Americas delivery rate is 99.4%.

Dotdigital, Email and SMS marketing benchmarks 2026 (EMEA and Americas editions)
98.05%

of the email Sinch Mailgun sent for Information Technology senders in 2025 was delivered, against a 0.14% bounce rate. Mailgun counts an email as delivered when the receiving server accepts it, and notes that an email delivered to the spam folder still counts as delivered.

Sinch Mailgun, The 2026 email impact report
98.4%

deliverability rate for Omnisend's e-commerce campaigns in 2025. Omnisend does not define the term or publish a bounce rate.

Omnisend, Email marketing benchmarks: What good looks like in 2026
Hard, soft or all bounces

Most bounces are soft. In Dotdigital's global data 0.74% of emails sent soft-bounced and 0.05% hard-bounced, a total of 0.79%; EMEA shows 0.79%, 0.06% and 0.84%. A hard-bounce rate and a total bounce rate are not comparable.

What the rate leaves out

Dotdigital's EMEA delivery rate (99.16%) and bounce rate (0.84%) add up to 100%. Mailgun's do not: Information Technology senders had 98.05% of email delivered and a 0.14% bounce rate in 2025, and the page does not say what happened to the rest.

Who sends through the platform

Mailgun's figures come from email sent through its sending service by its ten highest-volume industries (0.01% to 0.22%), Dotdigital's from marketing campaigns (0.75% to 0.84% in total), and Benchmark Email's from its users, by company size (1.05% to 5.74%).

Dated, undated and older data

Mailgun covers 2025, Dotdigital's 2026 posts give no data window, and Benchmark Email's page carries no date. GetResponse's report describes 2023 (see its report), and Mailchimp's benchmark page says it calculated soft and hard bounce rates by industry but publishes none.

What is a good email bounce rate?

It depends on how your platform counts and who you send to.

Total bounces (hard and soft)

Dotdigital's 2026 averages run from 0.75% (APAC) to 0.84% (EMEA), with a global average of 0.79% of emails sent. None of the sources we can quote sets a 'good' bounce-rate threshold from its own data, so compare with a figure that counts bounces the way your platform does.

Hard bounces

Dotdigital's hard bounce rates are 0.04% to 0.06% of emails sent (global 0.05%). Mailgun describes a hard bounce as an address that does not exist; soft bounces are temporary and make up most of Dotdigital's total.

By industry or company size

Mailgun's 2025 industry rates run from 0.01% (Air Freight & Logistics) to 0.22% (Professional Services), with Retail at 0.18%. Benchmark Email's undated averages by company size run from 1.05% to 5.74%. Neither says what its rate counts.

Before you compare

  1. Email bounce rate and website bounce rate are different metrics: one counts emails that could not be delivered, the other website visits that leave without engaging.
  2. Do not compare a hard-bounce rate with a total bounce rate: in Dotdigital's global data they are 0.05% and 0.79% for the same sends.
  3. Dotdigital's Americas page prints a total bounce rate of 0.06%, below its own soft bounce rate of 0.56%; soft plus hard bounces (0.60%) and its 99.4% delivery rate add up to 100%. We leave the 0.06% figure out.
Show 1 more cautionsShow fewer cautions
  1. Cold-outreach tools publish bounce rates for emails sent to people who never subscribed. That is a different population, so those figures are not on this page.

How to calculate email bounce rate

Bounce rate =bounced emails ÷ emails sent× 100
  • Divide by emails sent, not delivered: a bounced email was never delivered. Dotdigital's soft, hard and total bounce formulas all divide by emails sent.
  • Track hard and soft bounces separately. Mailgun describes hard bounces as addresses that do not exist and soft bounces as temporary failures (full inbox, server timeout, momentary block).
  • Delivery rate is delivered emails ÷ emails sent × 100. Dotdigital's pages print this formula the other way round.

Questions about email bounce rates

What is a good email bounce rate?

None of the sources we can quote sets a 'good' threshold from its own data, so compare with published figures that count bounces the same way. Dotdigital's total bounce rate (hard and soft) averages 0.79% of emails sent worldwide, from 0.75% in APAC to 0.84% in EMEA; hard bounces alone are 0.04% to 0.06%. Mailgun's 2025 industry figures are lower (0.01% to 0.22%) but do not say what they count.

What is the average email bounce rate?

There is no single average, and we do not blend the sources into one. Dotdigital's 2026 global average is 0.79% of emails sent. Mailgun's 2025 industry rates run from 0.01% to 0.22%, and Benchmark Email's undated company-size averages from 1.05% to 5.74%.

What is the difference between a hard bounce and a soft bounce?

A hard bounce is permanent: Mailgun describes it as an address that does not exist. A soft bounce is temporary, such as a full inbox, a server timeout or a momentary block. In Dotdigital's global data soft bounces (0.74% of emails sent) far outnumber hard bounces (0.05%).

Is a 50% bounce rate bad?

For email, 50% would be far above every figure on this page; the highest is Benchmark Email's undated 5.74% for companies with 100 to 250 employees. Check which metric the figure describes: website bounce rate counts website visits that leave without engaging, and its benchmarks are not comparable with email delivery failures.

Is bounce rate calculated on sent or delivered emails?

On emails sent: a bounced email was never delivered. Dotdigital's formulas divide soft, hard and total bounces by the number of emails sent. Mailgun and Benchmark Email do not publish a formula.

Tools to measure and improve it

Chosen for what they do, not ranked.

7.0 out of 10
NeverBounce

Removes invalid and risky addresses before a send, which protects deliverability and the rates you measure.

7.2 out of 10
Mailgun

Email sending API with a free tier of 100 emails a day; Mailgun also publishes the industry bounce and delivery rates on this page.

Freemium · $15/moRead review Mailgun
7.2 out of 10
Benchmark Email

Email platform with a free plan up to 500 contacts; Benchmark Email also publishes the company-size bounce rates on this page.

Freemium · ~$13/moRead review Benchmark Email
7.6 out of 10
Brevo

Brevo says its platform handles hard bounces automatically; priced by emails sent rather than list size.

Freemium · $9/moRead review Brevo
Email marketing software, compared
Sources and licence notes8 sources · 5 quoted · 2 linked only · checked Nov 12, 2026
  1. Dotdigital, Email and SMS marketing benchmarks 2026: Americas edition2026 edition (public regional post; full Global Benchmark Report is gated) · last checked Nov 12, 2026Website terms (updated 2019-01-16) assert ownership of site content but contain no consent requirement or ban on citing figures. Regional figures quoted with attribution.0.56%Grade B
  2. Dotdigital, Email and SMS marketing benchmarks 2026: EMEA edition2026 edition (companion regional post to the Americas edition) · last checked Nov 12, 2026Same Dotdigital website terms (silent on citing figures).0.84%Grade B
  3. Dotdigital, Email and SMS marketing benchmarks 2026: APAC edition2026 edition (companion regional post) · last checked Nov 12, 2026Same Dotdigital website terms (silent on citing figures).0.79%Grade B
  4. Sinch Mailgun, The 2026 email impact report: Email benchmarks: How do you measure up?2026 edition (2025 sending data) · last checked Nov 12, 2026Sinch Email Terms of Service (last revised 2025-02-15) apply to Services ordered under a Service Order; they contain no clause on website content, copying or citing, and only the name and logo need prior written consent (s.12). Silent on citing: a few figures, attributed and linked.0.22%Grade B
  5. Benchmark Email, Email Marketing Benchmarks By Industry and RegionUndated · last checked Nov 12, 2026Terms of use (updated 2026-07-10) cover the platform; silent on citing website content. Overall figure quoted with attribution.5.74%Grade B
  6. Omnisend, Email marketing benchmarks: What good looks like in 20262026 article (2025 data) · last checked Nov 12, 2026Same Omnisend terms as the report (restriction targets 'Omnisend Content' provided as part of the Services). A few figures, attributed and linked.Grade B
  7. GetResponse, 2024 Email Marketing Benchmarks2024 edition (no newer edition found) · last checked Nov 12, 2026Linked only. Confirmed link_only: terms (effective 2026-06-09, §XIII) require express written consent to reproduce 'GetResponse Property', defined to include content on the Site and Websites.Grade A
  8. Validity, 2026 Email Deliverability Benchmark Report2026 edition (2025 and earlier editions listed) · last checked Nov 12, 2026Linked only. Terms of Service (last updated 2025-09-25) govern the Services and say nothing about website content, but the report sits behind a form and the landing page shows no figures, so nothing can be shown. Link only.Grade B

These are figures reported by the platforms themselves about their own customers; we do not measure them ourselves. Found an error or a newer edition? Tell us

Related