Updated · 8 min read
Bounce rate management: the thresholds and the fix order
Picture the dashboard you check on Monday morning. There's a bounce chart on it. The number sits below 2%, you nod, you move on. That's the trap. Bounce rate is a metric with texture — hard versus soft, by ISP, by cohort — and the texture is where every operational decision lives. Skip the texture and you'll either panic at noise or sleepwalk past a real problem until your reputation's already in the hole.
By Justin Williames
Founder, Orbit · 10+ years in lifecycle marketing
The two kinds of bounce, and why one is fatal
Bounce rate is the fastest-moving deliverability signal you have. Complaint rate — the percentage of recipients who hit "mark as spam" — is slower but nastier when it shifts. By the time complaints move, the reputation is already in the hole.
Every email you send ends in one of three places: the inbox, the spam folder, or back at your sending server with an error attached. That error is a bounce — the receiving mail server saying "no, not delivering this." And the receiving server tells you why, in two flavours that determine everything you do next.
Hard bounces are permanent. The address doesn't exist, has been disabled, or is otherwise unreachable forever. Treatment is non-negotiable: immediate permanent suppression — adding the address to a do-not-send list — covered in detail in the list hygiene policy. No retries. Retrying a hard bounce is one of the cheapest ways to damage sender reputation— the score mailbox providers (Gmail, Outlook, Yahoo) keep on your sending domain to decide whether to trust you. Every retry is a permanent negative signal you're paying to re-send.
Soft bounces are temporary. Mailbox full. Server briefly unavailable. Greylisting — when a receiving server defers you on first contact and asks you to try again, a basic spam-filter trick. Retry a reasonable number of times, then suppress. Three consecutive failures is the operator default; some teams stretch to five.
Here's the gotcha that catches people. Some providers report soft bounces that are actually persistent. A mailbox that's been "temporarily full" for six months isn't getting emptied — the user has abandoned the address. When an address has soft-bounced on every send for three months or more, treat it as hard regardless of what category the provider tagged it with. Trust the pattern over the label.
What "normal" looks like — and where the cliff is
Before the thresholds, the mental model: bounce rate is a rolling number, not a per-send one. ISPs — internet service providers, the companies running the mail servers your audience uses — judge you on a 30-day moving average, not on yesterday's send. One bad broadcast won't sink you if the rest of the program is healthy. A bad fortnight will.
< 2%
Healthy total bounce rate. Below this, no action needed.
2–5%
Warning zone. Investigate acquisition channels; audit recent imports.
> 5%
ISPs start rate-limiting. Gmail and Yahoo may temporarily block further sending.
The cliff at 5% is real. Once Gmail or Yahoo decide your sending pattern looks careless, they don't just penalise the bouncing addresses — they slow down or block your delivery to the engaged ones too. Everything else gets harder at exactly the moment you're trying to fix the underlying problem. That's the asymmetry to remember: bounce rate doesn't just measure a problem, it actively makes the next problem worse.
The deliverability guide covers the broader connection between bounce rate and reputation. The headline: bounce rate is the early warning. Complaint rate is the funeral.
Why bounce rates climb — five causes, ranked by how often we see them
When the chart starts climbing, it's one of five things. Almost always. They're ranked here by frequency, which is also the order to walk through during diagnosis — start at the top, only move down when the symptoms don't fit.
1. Stale list data.The single most common cause. Addresses that were valid eighteen months ago have been abandoned, disabled, or rotated as people changed jobs. Fix with an aggressive sunset policy — automatic suppression of anyone who hasn't opened or clicked in a defined window — plus list validation (a third-party check that pings each address to see if it still exists) on any long-dormant segment before sending to it. Shape: a gradual climb, not a sudden spike.
2. Bad acquisition.An import from a legacy source. A promotion that accepted fake emails. A co-marketing deal where the partner's list was dirty and nobody audited before importing. Trace bounces back to the acquisition path — high-bounce segments almost always trace to a specific event you can name and isolate.
3. Authentication failure. SPF, DKIM, and DMARC — the three DNS records that prove a sending domain is allowed to send mail and that the message wasn't tampered with — suddenly failing. Bounces climb because receiving servers reject what they can't authenticate. Shape: a sharp spike, not a gradual rise. Check DMARC reports, confirm DNS hasn't changed, verify the configuration on your ESP (email service provider — Braze, Iterable, Klaviyo, the platform that actually does the sending). Covered in the authentication guide.
4. Reputation damage.Accumulated negative signal — volume spike, complaint surge, authentication issues — and one or more ISPs are now blocking you specifically. Shape: bounces concentrated at a single provider. Gmail and nowhere else. Microsoft and nowhere else. Recovery is the only path: scale volume back, segment tightly to engaged users, monitor through Postmaster Tools (Google's reputation dashboard for senders) and SNDS (Microsoft's equivalent — Smart Network Data Services).
5. Content triggering spam filters.Specific message content tripping a provider's filter — usually a link to a flagged domain, a suspicious attachment pattern, or phrasing that pattern-matches to a recent scam wave. Shape: bounces on one campaign, not across the program. Isolate the campaign, run a seed test (send to a small list of test addresses across providers and see where it lands), adjust the content.
When the chart spikes — what to do in the first thirty minutes
1. Stop and segment. Pause non-critical sends. Bucket the bounces by ISP, by campaign, by cohort. Concentrated at one ISP means a reputation problem with that ISP. Spread evenly across all of them means authentication or list data. The shape of the spike does most of the diagnostic work for you.
2. Apply immediate suppression. Hard bounces from the spike go to permanent suppression now, not after the post-mortem. Soft bounces past the retry threshold go to temporary suppression. Stop the bleeding before you start the surgery.
3. Investigate the pattern. Walk the five causes in order — list, acquisition, authentication, reputation, content. The shape of the spike tells you more than any single metric. A gradual climb at all ISPs is list rot. A sharp spike across all ISPs is auth. Concentration at one ISP is reputation. Let the symptoms route you.
4. Fix the underlying cause. Authentication is fast — minutes to hours once diagnosed. Reputation recovery is slow — three to six weeks for most programs, and there's no shortcut. List-data fixes sit in the middle: once the bad acquisition path is found, the cleanup itself is straightforward.
5. Resume gradually. Don't snap back to full volume the day the fix deploys. Ramp over a week, monitoring bounce rate daily. Impatience here is how a one-off incident becomes a recurring pattern.
The Orbit Deliverability Management skill runs this triage and produces a structured diagnosis. Worth wiring in before a spike happens, not during one — incident pressure is exactly the condition under which people skip step one.
Read to the end
Scroll to the bottom of the guide — we'll tick it on your reading path automatically.
Frequently asked questions
- What's a good email bounce rate?
- Hard-bounce rate below 1% is healthy. 1-2% is the warning zone — usually signals list-acquisition problems (purchased lists, scraped emails, typo traps). Above 2% is a reputation risk and mailbox providers start rate-limiting. Soft-bounce rate has less strict thresholds — 3-5% can be normal depending on your audience's behaviour, but chronic soft bounces on the same addresses should convert to suppression after 14 days.
- How do I reduce my email bounce rate?
- Three steps. First, fix list acquisition — remove any third-party lists, add real-time email verification to signup forms (ZeroBounce, NeverBounce, similar), require double opt-in for uncertain sources. Second, audit the existing list — run it through a verification service, suppress hard bounces immediately, and set up soft-to-hard conversion at 7-14 days. Third, watch the trend — bounce rate should decline after interventions. If it doesn't, the problem is acquisition, not hygiene.
- Why do hard bounces damage sender reputation?
- Mailbox providers interpret repeated sends to invalid addresses as a sign the sender has poor list hygiene, doesn't honour unsubscribes, or is using scraped / purchased lists. The reputation hit is substantial — a campaign to a list with 5% hard-bounce rate can drop domain reputation visibly in Google Postmaster Tools within 24-48 hours. Recovery takes weeks of clean sending. The discipline: every hard bounce should be in suppression before the next campaign sends to that address.
- What's the difference between a bounce and a block?
- A bounce is a delivery failure reported by the receiving mail server — your ESP accepted the message, attempted delivery, and got an error back. A block is the receiving server refusing the connection upfront (or accepting and dropping silently). Blocks often don't produce bounce records in your ESP because the message never reached the recipient's server. Bounce rate reporting can look deceptively clean even when blocks are high — only seed-list testing + Google Postmaster reveal blocks.
- Should I send to a list I haven't mailed in six months?
- Not without a re-engagement sequence first. A list that's been dormant for six months has accumulated hard bounces silently, users have forgotten they subscribed, and a cold send to it triggers reputation damage from the mass complaints and bounces. Correct approach: run a reconfirmation or winback sequence to the most-recently-engaged subset first, prove reputation is intact, then gradually extend to the full dormant cohort over several weeks.
This guide is backed by an Orbit skill
Related guides
Browse allList hygiene: the six-rule policy
List hygiene isn't cleanup; it's a continuous policy that runs automatically. Here's the six-rule policy every lifecycle program should have written down, each tied to a specific deliverability outcome.
Spam complaints: the playbook for detecting and reducing them
Spam complaints are the hardest-hitting negative reputation signal in email. They compound faster than bounces and recover slower. This is the playbook — what actually triggers them, how to catch them early, and the four levers that reliably bring the rate back down.
Bounces vs blocks vs deferrals: what your ESP's error codes actually mean
SMTP error codes are the mechanical reason emails don't arrive. Most programs lump them all into 'bounces' and miss the signal. Here's the actual taxonomy — hard bounce, soft bounce, block, deferral — and what each one tells you.
Reputation recovery: the 90-day playbook for dropping from High to Low
Domain reputation at Low or Bad isn't a problem you fix this week. It's a 6–12 week project of disciplined sending to engaged users only, while the reputation signals slowly reset. Here's the plan.
Email deliverability — the practitioner's guide
Deliverability isn't a setting. It's the running total of every send decision you've made since you bought the domain. Four pillars hold it up. Break one and the whole program starts leaking.
The unsubscribe page is the most important page in your lifecycle program
The page every lifecycle team ignores is the one quietly deciding sender reputation, suppression-list quality, and the fate of next quarter's deliverability. A short defence of why it deserves the ten-minute rebuild.
Found this useful? Share it with your team.
Use this in Claude
Run this methodology inside your Claude sessions.
Orbit turns every guide on this site into an executable Claude skill — 63 lifecycle methodologies, 91 MCP tools, native Braze integration. Free for everyone.