How to Reduce Your Email Bounce Rate to Under 2%

I run cold email lists for a living. Elevate Clients Inc. is a B2B agency, and the reason this site exists is that verification kept being the step that decided whether a campaign worked. Bounce rate is the number that tells you whether you got that step right, and it is the number most guides treat as a content problem when it is overwhelmingly a data problem.

What bounce rate is

Bounce rate is the share of your sent messages that a receiving mail server refused to deliver. It is expressed as a percentage of messages that reached a server and got rejected, which is not the same as a percentage of the list you started with. Addresses removed before sending never had the chance to bounce and should not appear in the denominator, though plenty of tools quietly include them and flatter the figure.

Hard bounces and soft bounces

The distinction matters because only one of them damages you.

A hard bounce is permanent. The mailbox does not exist, or the domain cannot receive mail at all. There is no version of the future in which retrying helps, and these are the bounces that providers read as a signal about the sender. Remove them immediately and add them to a suppression list so they cannot re-enter through a later import.

A soft bounce is temporary: a mailbox over quota, a server briefly unavailable, a message refused for size. The address may work perfectly well next week. Retry a small number of times, and if it keeps failing across sends, treat it as hard and suppress it.

What counts as a good bounce rate

The convention in deliverability work is that cold outreach should sit under 2 percent, and permission-based sending to an engaged list should sit under half a percent. It is worth being precise about where those figures come from: mailbox providers do not publish the thresholds they act on. The numbers are inferred from observed behaviour and passed around as working rules.

Treat them accordingly. Under 2 percent is not a limit you can safely sit just beneath, it is the point past which you should assume something is wrong with the list rather than the copy. Sustained rates above 5 percent warrant pausing the campaign rather than adjusting it.

Bounce rate is not a measure of your writing. It is a measure of how well you know your list.

What bounce rate hides

Bounce rate is the best single indicator of list quality, and it is also incomplete in a way that matters for cold outreach.

It only counts messages a server refused. Mail that is accepted and then discarded never appears in it. That happens constantly on catch-all domains, where the server takes delivery of anything and decides internally what to do with it, and it happens with spam traps, which accept quietly by design. A campaign hitting either can post a flattering bounce rate while doing real damage to the domain sending it.

The practical consequence is that a low bounce rate on an unverified list is not evidence of a healthy list. It may simply mean a large share of the addresses sit behind servers that accept everything. If your reply rate is poor while your bounce rate looks excellent, that combination is worth investigating before you rewrite the copy again.

The four levers

Everything that moves bounce rate falls into four categories. They are not equally weighted: the first one accounts for most of the variance, and it is the one usually given a single bullet point.

1. List verification. Whether the addresses can receive mail at all.
2. Sending infrastructure. Whether receiving servers can confirm you are who you claim to be.
3. Warmup. Whether your sending pattern looks like a real correspondent rather than a sudden burst.
4. Ongoing hygiene. Whether the first three stay true over months rather than on one day.

Lever 1: list verification

Hard bounces are caused by sending to addresses that do not exist. Everything else on this page is secondary, because no amount of authentication or warmup makes a nonexistent mailbox accept mail.

The complication is that verification is not a single quality of service. Any verifier will remove the obviously dead addresses: malformed strings, lapsed domains, domains with no mail servers. Where they diverge is the segment that comes back ambiguous. On a typical B2B list, 25 to 30 percent of addresses sit on catch-all domains that accept mail for every address they are asked about, which makes the standard mailbox check inconclusive. Most tools report that condition and stop.

What that costs on a real list

Take a 10,000-address B2B list. A standard verification pass settles about 7,500 addresses cleanly and returns 2,500 as catch-all. You now have three choices and each one has a bounce-rate consequence.

Delete the 2,500 and your bounce rate will be excellent, because you have removed a quarter of your list including the addresses that were fine. Send to all 2,500 unverified and your bounce rate becomes a lottery, since some of those mailboxes do not exist and you have no way to know which. Or resolve the segment: a second verification pass returns a definitive verdict for roughly 72 percent of it, about 1,800 addresses, which is 18 percent of the whole list moving from unusable to confidently sendable.

That third option is the one that gets both numbers right at once, and it is the practical reason verification depth matters more than the headline accuracy figure any vendor quotes. How verification works covers the stages in detail, and catch-all domains covers why that segment behaves the way it does.

The cost side

Verification runs from $0.0070 per address at the smallest pack down to $0.0019 at the largest. On a 10,000-address list that is $55. Set that against what you paid to source those contacts, and against the weeks of degraded delivery that follow a domain being marked as a careless sender, and the decision stops being close. Verification is among the cheapest line items in an outreach budget and the one with the most direct effect on whether the rest of it works.

Lever 2: sending infrastructure

Infrastructure will not fix bounces caused by dead addresses, but bad infrastructure creates rejections that look like bounces in your reporting and damage you the same way.

Authentication. Three records need to exist and agree. SPF declares which servers may send for your domain. DKIM signs your messages so a receiver can confirm they were not altered and genuinely came from you. DMARC tells receivers what to do when those checks fail and, importantly, gives you reporting on what is being sent in your name. Missing or contradictory records get messages refused outright by strict receivers.

Move DMARC forward deliberately. A DMARC record starts at a policy of none, which asks receivers to report but not to act. That is the right place to begin, because it shows you what is being sent in your name before you start blocking anything. Read the reports for a few weeks, confirm every legitimate sender is passing, then tighten to quarantine and eventually to reject. Skipping straight to reject with an incomplete SPF record is a reliable way to have your own mail refused.

Separate your sending domains. Run cold outreach from domains dedicated to it, not from the domain your customers use to reach you. Reputation attaches to the domain, so this contains the damage when something goes wrong. It also means a problem on one outreach domain does not take your invoices and support replies down with it.

Keep volume per mailbox modest. A small number of messages per mailbox per day, spread across several mailboxes, looks far more like human correspondence than the same total from one address. This is a pattern question rather than a technical one, and receivers are demonstrably sensitive to it.

Lever 3: warmup

A brand new domain with no sending history that suddenly emits hundreds of messages is indistinguishable, from the receiver’s side, from a domain registered specifically to spam. Warmup is the practice of building a sending history before you need it.

The mechanics are straightforward. Start at a very low daily volume from each new mailbox, on the order of a handful of messages. Increase gradually over several weeks rather than days, and keep the increments small enough that no single day looks like a step change. Aim to generate genuine replies during this period, because engagement is the strongest positive signal available and its absence is conspicuous. Most teams treat two to four weeks as the minimum before a new domain carries real campaign volume.

Automated warmup services exist and are widely used. They work by exchanging mail between accounts in a network to manufacture the engagement signal. Whether that is worth doing is a judgement call about your own risk tolerance, and it is not a substitute for the gradual ramp itself. What matters more than the tooling is that the curve is smooth and the domain has history before it matters.

Lever 4: ongoing hygiene

The first three levers describe a state. This one describes keeping it, which is where most teams lose the number they worked to get.

• Re-verify on a schedule. Before every campaign for cold lists, quarterly for lists that generate engagement. Address decay is continuous and invisible until you send.
• Maintain a real suppression list. Every hard bounce goes on it permanently. The most common cause of a bounce rate creeping back up is a re-import of an old file containing addresses you already proved were dead.
• Monitor per domain, not in aggregate. Reputation attaches to individual sending domains, so one failing domain hides comfortably inside a healthy average until it is doing real damage.
• Watch the trend. A single elevated send is noise. Three in a row moving the same direction is a signal, and it is usually telling you the list has aged.

Watch the signals bounce rate does not cover

Bounce rate is one input and a lagging one. Three others tell you about deliverability earlier, and all three are available without extra tooling. Spam complaint rate is the most serious: complaints are read as a much stronger negative signal than bounces, and the tolerance for them is far lower. Reply rate is the best available proxy for whether messages are reaching inboxes rather than spam folders, because a sudden fall with unchanged copy usually means placement has moved rather than the writing has. And open behaviour, treated cautiously given how unreliable open tracking has become, is still useful as a relative measure across your own sends.

Why bounce rates get high

In order of how often they turn out to be the culprit:

• The list was never verified. By far the most common, and the easiest to fix.
• The list is old. Contacts collected a year ago and never re-checked. People move, and a startup list ages faster than an enterprise one.
• The data was scraped or guessed. Pattern-generated addresses of the first.last@company.com variety are guesses, and a meaningful share of them are wrong by construction.
• Generic addresses dominate. Role accounts such as info@ and support@ are usually real, so they do not bounce, but they attract filtering and engage poorly. They inflate the problems bounce rate is a proxy for without inflating bounce rate itself.
• A merge of two lists reintroduced dead addresses. Combining a cleaned export with an older file quietly restores everything you removed. This is the most frustrating cause because the list was verified, correctly, before somebody appended to it. Suppression lists exist precisely to survive this.

Fixing a high bounce rate now

If a campaign is currently running hot, this is the order that stops the damage fastest.

1. Pause the campaign. Every additional send while the rate is elevated adds to the reputation cost. This is the step people resist and the one that matters most.
2. Re-verify the remaining list. Not the portion you have already sent to, the portion you have not. This is where the preventable bounces still are.
3. Remove and suppress every undeliverable address. Permanently, in the source system rather than only in the sending tool.
4. Segment the ambiguous addresses. Catch-all and risky results go into their own segment. Resolve them with a second pass, or hold them back until you can, but do not leave them mixed into the main list.
5. Resume at reduced volume. Restart below your previous daily send and increase over a week or two while watching the rate. Treat it as a partial warmup, because from the receiver’s perspective that is what it is.

The verification steps above are what our two-pass engine is built for, and pricing is credit-based with no subscription, so cleaning one list does not commit you to anything ongoing. If you want to test the theory on the segment that is actually causing your bounces, a free account includes 100 credits.

Frequently asked questions