Definition
Email bounce rate is the percentage of messages you sent that the receiving mail server refused to accept. The recipient never sees the email. Instead, the receiving server sends back a bounce notification with an SMTP response code explaining why. Bounces come in two flavors. Hard bounces are permanent failures, usually because the address is invalid, the domain doesn't exist, or the receiving server has decided not to accept your mail at all. Soft bounces are temporary, usually because the recipient's mailbox is full, the receiving server is overloaded, or there's a brief authentication or rate-limit issue.
How it's calculated
Simple formula: bounces ÷ sent × 100. The denominator matters. Some platforms calculate against "delivered" instead of "sent", which makes the number look better than it is. The honest version uses sent as the denominator, because a bounce is a failure to deliver, and you want to know what fraction of your attempts failed.
Most reporting tools split bounce rate into hard bounce rate and soft bounce rate because the two need different responses. Hard bounces should immediately suppress the address from future sends. Soft bounces are retried with backoff, and only suppressed if they continue to fail over a window of days.
Target thresholds
Industry benchmarks (vary slightly by source and segment):
- Hard bounce rate below 0.5%. Below 0.2% is excellent. Above 1% starts hurting your sender reputation. Above 5% is usually a hard ceiling that triggers mailbox provider intervention.
- Total bounce rate below 2%. Includes soft bounces. Above 5% suggests systemic problems beyond just list hygiene.
- For transactional email specifically, hard bounce under 0.2%. The audience is signed-up users whose addresses you collected directly. There's no excuse for high hard bounce here.
The mailbox providers don't publish their exact thresholds, but Gmail's documentation hints at 0.3% as a danger zone for spam complaints (related, not bounces, but the dynamic is similar). Outlook tends to be stricter than Gmail on bounce-rate signals.
What causes a high bounce rate
List quality problems (the most common cause)
Buying or scraping a list is the fastest way to a 5%+ hard bounce rate. Old lists you haven't mailed in 12+ months degrade by roughly 25-30% per year as people change jobs and abandon addresses. Even good lists collected through legitimate signup forms accumulate bad addresses if you don't have a confirmed opt-in (double opt-in) flow, because typos go undetected.
Authentication or reputation problems
If SPF, DKIM, or DMARC are misconfigured, some receivers will hard-bounce the message rather than accept and risk policy violations. This shows up as bounces, but the real problem is the auth chain. Same for reputation: if your domain or IP reputation is bad enough, receivers will reject at the SMTP edge, which counts as a hard bounce in most reporting.
Content and structure problems (less common)
Some receivers reject for content reasons: malformed headers, oversized messages, suspicious attachments, broken DKIM signatures from message tampering somewhere in the pipeline. These show up as bounces but the fix is in your sending infrastructure, not the address list.
How to bring it down
In order of impact:
- Suppress hard bounces automatically. Once an address hard-bounces once, never send to it again. Every additional attempt is a vote against your reputation.
- Implement double opt-in on signup forms. Cuts typo-driven bounces by 80-90%. Worth the small drop in signup completion rate.
- Re-engage or remove dormant subscribers. If they haven't opened anything in 6 months, send a re-engagement campaign. If they still don't engage, suppress them.
- Verify your list before a big send. Email verification services (Kickbox, NeverBounce, etc.) catch invalid addresses before you send to them, in exchange for a per-address fee. Worth it before any high-volume push.
- Fix authentication. Publish SPF, valid DKIM, a DMARC policy. If you can't get this right, you can't fix bounce rate without it.
- Don't send to role addresses unnecessarily. info@, support@, admin@ tend to bounce more and complain more.
How sendmsg.io handles bounces
The Cortex engine classifies every bounce in real time as it lands in our Postfix logs. Hard bounces hit a suppression list automatically and never re-send. Soft bounces follow an exponential backoff retry policy: first retry in minutes, second in an hour, then a few hours, with a final attempt 24 hours out. If the soft bounce continues to fail past that window, the address gets suppressed pending a manual override.
We also surface bounce rate as a per-domain temperature signal in the dashboard. If a customer's bounce rate crosses 1% in a rolling window, we apply graduated throttle controls to that domain before mailbox providers do it for us at a much higher cost. The principle is the same as with sender reputation: catch the slope before it becomes the cliff.
Related reading
- Sender Reputation: how bounce rate feeds into the score
- Email Deliverability: the outcome bounce rate affects
- DKIM, SPF, DMARC: the auth trio that prevents auth-driven bounces
- Email Reputation Management: how sendmsg.io's Cortex protects you
- Full deliverability guide