Email bounce rate: the math, the thresholds and how to lower it

Send a campaign to 10,000 addresses and get 220 failure notices back, and your bounce rate for that send is 2.2% — 220 divided by 10,000. The arithmetic is the easy part. What makes bounce rate worth a full guide is everything the single number hides: two very different failure types averaged together, platform enforcement thresholds that can pause your account, and a reputation effect at mailbox providers that outlasts the campaign that caused it.

What the number actually measures

Bounce rate is bounced messages divided by messages attempted, per campaign or per period. Both terms need unpacking.

The numerator mixes two categories with opposite meanings. Hard bounces are permanent failures — the address or domain doesn't exist, or the receiver refuses it outright. Soft bounces are temporary (full mailbox, busy server, oversized message) and most resolve on retry. A 2% rate composed of hard bounces says your list contains dead addresses; the same 2% in soft bounces during a receiver outage says almost nothing. Any bounce-rate analysis that doesn't split the two is reading a blended average of a signal and its noise. Which failures belong to which class, and what to do about each, is the subject of our hard bounce vs soft bounce decision guide.

The denominator varies by platform: some divide by messages sent, others by messages sent minus prior suppressions. And platforms differ in when they count a chronic soft bounce as hard. The practical consequence — bounce rates aren't precisely comparable across platforms, so track your own trend rather than benchmarking your number against a friend's on different software.

Where the thresholds sit

No standards body defines an acceptable bounce rate, so the numbers in circulation come from two places worth distinguishing.

The industry convention puts the comfort line at about 2% per campaign, with well-maintained lists typically running under 1%. That figure is a practitioner rule of thumb, not a measurement — we cite it as the convention it is. Its logic is sound, though: organic list decay between regular sends produces some baseline of newly-dead addresses, and rates meaningfully above that baseline imply collection or hygiene problems rather than natural churn.

Platform enforcement thresholds are the ones with teeth. ESPs police bounce rates because high-bouncing customers damage shared infrastructure — one tenant's dirty list degrades the shared IPs every tenant sends from. Most platforms publish acceptable-use limits and will warn, pause campaigns or suspend accounts that exceed them; the specific numbers are each platform's own published policy, so check yours before you need to know it.

Mailbox providers, meanwhile, publish no bounce-rate threshold at all — but repeated delivery attempts to nonexistent addresses are a classic input to sender reputation, because a sender who doesn't know which of their addresses exist looks like a sender who never had permission in the first place. That's the pattern of purchased lists, and it travels with a worse companion: lists dirty enough to bounce heavily are the lists that contain spam traps, which never bounce and never forgive.

Why bounce rate moves

Sudden spikes and slow drifts have different cause lists.

A spike on one campaign usually means: an old segment or imported file entered the send, a receiver-side event (one domain deferring or rejecting everything — check whether bounces cluster by domain), or an authentication or blocklist problem being returned as bounces. Read the bounce messages before touching the list; text citing policy, blocklists or SPF/DKIM points at your sending, not your subscribers, and the fix runs through the email authentication checker and blacklist checker rather than through deletions.

One timing wrinkle worth knowing: not every bounce arrives during the send. A receiver can accept a message and generate a failure notice minutes or hours later — an asynchronous bounce — which is why a campaign's bounce count can keep climbing after the send completes, and why reading the rate five minutes in undercounts it. Judge the number once the send has aged a day.

A slow drift upward is list decay outpacing hygiene: addresses dying between sends faster than engagement filtering retires them. The drift accelerates with infrequent sending — a list mailed quarterly delivers three months of accumulated deaths per campaign, which is one of several reasons irregular blast patterns underperform steady cadence.

Lowering it: the order that works

1. Stop new bad addresses at the door. Collection quality dominates everything downstream. Double opt-in confirms each address can receive and act on mail before it enters circulation; form-level validation catches typos at the moment the subscriber can still fix them.
2. Let suppression do its job. Hard bounces go to the suppression list automatically on any serious platform. The rate-killer is re-importing stale exports over that history — treat every list import as a hazardous operation.
3. Retire the silent before they die. A standing list hygiene policy — re-permission at the edge of a defined inactivity window, suppression after — removes addresses in the abandoned-but-not-yet-closed phase, which is where tomorrow's hard bounces live today.
4. Escalate chronic soft bounces. An address soft-bouncing across several consecutive campaigns is permanent failure in temporary clothing; convert it to suppressed if your platform doesn't already.
5. Be cautious with bulk verification services. Pre-send verification of an old or imported list can trim obvious dead addresses, but verifiers can't detect traps or lapsed consent — accuracy claims are vendor claims, and a verified purchased list is still a purchased list. Verification narrows a bounce problem; only consent and hygiene solve one.

Monitoring it properly

Three habits turn bounce rate from a post-mortem number into an early-warning system. Split hard from soft in whatever report you watch — the blend hides the signal. Watch the trend across campaigns, not the single-send value, since one receiver outage can distort any individual send. And when the number moves, read actual bounce texts before acting; the classification labels triage, but the text diagnoses, and the difference decides whether you should be cleaning your list or fixing your DNS. Bounce rate is one input in the wider metric set — where it sits among complaint rate, engagement and placement is mapped in what email deliverability is.

Related guides

• Hard bounce vs soft bounce — the decision guide for individual bounces
• Hard bounce — permanent failures, defined
• Soft bounce — temporary failures and retry mechanics
• List hygiene — the practice that keeps the rate down
• Suppression list — where bounced addresses must stay
• Email authentication checker — when bounce text blames your setup instead of the address

About this guide

Written by InboxRatio Editorial. The 2% comfort line is labeled as the industry convention it is; platform enforcement thresholds are each provider's own published policy and are cited as such. No vendor sponsorship influences this guide.

Methodology

InboxRatio's benchmark sends real campaigns through the platforms we review to controlled seed lists and records delivery outcomes, including how platforms classify bounces. The guidance here rests on platform documentation and labeled conventions rather than our measurements; the testing protocol is documented in how we test, and source rules in sources.

Last updated

5 September 2026. Convention and enforcement guidance reviewed quarterly.

Frequently asked questions

What is a good email bounce rate? The working convention: under 2% per campaign is acceptable, under 1% is where well-maintained lists run, and anything above 5% signals a list-quality problem needing intervention. These are industry rules of thumb — your platform's published enforcement thresholds are the numbers with consequences.

How is email bounce rate calculated? Bounced messages divided by messages attempted, per campaign: 220 bounces on 10,000 sent is 2.2%. Platforms differ slightly in denominators and in when chronic soft bounces get reclassified, so compare your trend over time rather than absolute numbers across platforms.

Do soft bounces count toward bounce rate? In most platforms' reported rate, yes — which is why splitting the number matters. A soft-bounce-heavy rate during a receiver outage is noise; the same rate in hard bounces is a list problem.

Does a high bounce rate hurt deliverability? Yes, twice over. Mailbox providers read repeated sends to dead addresses as evidence of poor list acquisition, degrading sender reputation. And your ESP enforces its own limits — sustained high rates draw warnings, campaign pauses or suspension.

How do I reduce my email bounce rate? In order of leverage: confirmed opt-in at collection, respect your suppression list (especially across imports and platform moves), a standing inactivity sunset policy, and escalation of chronic soft bounces. Verification services can triage an old list but don't substitute for consent-based collection.

Why did my bounce rate suddenly spike? Check three things in order: whether an old segment or fresh import entered the send, whether bounces cluster on one receiving domain (a receiver-side event), and what the bounce texts say — policy or authentication language means the problem is your sending setup, not the addresses.

Split your last campaign's bounces into hard and soft today and read ten of the actual bounce messages — then, if any of them mention policy or authentication, spend the next two minutes in the email authentication checker before you delete a single address.