Should You Email Catch-All Addresses? A Decision Framework

Send to catch-all addresses selectively, not all or nothing. Score each address by where it came from, send a test batch of 100 to 300 from the best-scoring group, measure the bounce rate, then send only as many more as your overall bounce limit can absorb. The formula and a worked example are below.

If you need the background on why these addresses cannot be verified, read what is a catch-all domain first. This post assumes you already have a list with a catch-all or accept-all segment and want to decide what to do with it.

The decision in one table

Catch-all addresses differ enormously in quality, and the biggest predictor is where the address came from. No verifier can tell you whether a specific catch-all mailbox exists, but you usually know how you got it.

Source of the address	Risk	Default action
Published on the company's own website, with the person's name	Low	Send in the test batch
Given to you directly (business card, form fill, reply from them)	Low	Send in the test batch
Found by a finder tool and matches a pattern confirmed elsewhere at that company	Medium	Send after the test batch passes
Guessed from a common pattern, no confirmation	High	Send only if the test batch comes back clean, and in small numbers
From a purchased, scraped or very old list	Very high	Do not send
A role address (info@, sales@) on a catch-all domain	Depends on use case	See should you send to role addresses

Two of these need a word of explanation.

A pattern confirmed elsewhere means you have at least one verified or published address at the same company showing its format, for example a press contact listed as jane.doe@company.com. A guess of john.smith@company.com then has evidence behind it. Our guide to company email formats covers how to establish this.

Purchased lists are excluded not because catch-all is worse there, but because you have no idea how old the addresses are, and catch-all domains will silently accept addresses that died years ago until the company tightens its configuration. See should you buy an email list.

Step 1: know your bounce limit

Your limit is set by whoever carries your mail, and you want to stay well under it. Published examples:

• Amazon SES says in its reputation metrics documentation: "If your bounce rate is 5% or greater, we automatically place your account under review. If your bounce rate is 10% or greater, we might pause your account's ability to send additional email". It recommends staying "below 5%".
• Clearout describes its own "risky" category as "generally usable when the bounce rate must remain below 5%" in its verifier documentation.

Other platforms have their own thresholds, collected in bounce rate limits by platform. For cold email from your own inboxes there is no published threshold at all, which is a reason for more caution, not less.

For the worked example below we use a target of 2% blended bounce rate, which leaves a wide margin under a 5% review line.

Step 2: measure, do not guess, with a test batch

The only way to learn a catch-all segment's bounce rate is to send to part of it and count. The size of that test batch decides how much the result tells you.

When a test batch has no bounces, the rule of three gives a quick upper limit: with zero events in n trials, you can be 95% confident the true rate is below 3/n.

Test batch size	Bounces	What you can say (95% confidence)
30	0	True rate below 10%. Nearly useless
100	0	Below 3% (Wilson interval gives 3.7%)
300	0	Below 1% (Wilson interval gives 1.3%)
100	3	Somewhere between about 1% and 8.5%
200	10	Somewhere between about 2.7% and 9.0%
200	24	Somewhere between about 8.2% and 17.2%

The ranges with bounces in them are Wilson score intervals, which we calculated for this post. The point of the table is the width: a test of 30 tells you almost nothing, and even 200 sends leave a range several points wide. Plan on the top of the range, not the middle.

Pick the test batch from your best-scoring group in the table above. If that group bounces badly, the lower groups will be worse.

Step 3: cap the catch-all volume with one formula

Once you have a measured catch-all bounce rate, you can calculate how many you can send without breaking your target. You need four numbers:

• V: how many confirmed-valid addresses are in the same send
• b_v: the bounce rate you expect from valid addresses (from your own past sends)
• b_c: the catch-all bounce rate from your test batch (use the top of the range)
• T: your target blended bounce rate

The maximum number of catch-all addresses, C, is:

C = V × (T − b_v) / (b_c − T)

It only applies when b_c is above T. If your catch-all segment bounces below your target, it is not the problem.

Worked example

These numbers are illustrative assumptions, not measurements. Substitute your own.

You have 8,000 confirmed-valid addresses and 2,000 catch-all addresses. Your valid addresses have historically bounced at 0.5%. Your target is 2%.

Measured catch-all bounce rate (top of range)	Maximum catch-all you can add	Blended rate if you sent all 2,000
5%	8,000 × 0.015 / 0.03 = 4,000	1.4%
9%	8,000 × 0.015 / 0.07 = about 1,710	2.2%
12%	8,000 × 0.015 / 0.10 = 1,200	2.8%

At 5%, the whole segment fits comfortably. At 9%, you can send about 1,700 of the 2,000, so drop the lowest-scoring 300. At 12%, send 1,200 and hold the rest, or go back and improve their sourcing.

Notice what drives the answer: the size of your valid segment. Catch-all addresses are only affordable when they ride alongside confirmed ones. Sending a batch that is mostly catch-all, with few confirmed addresses to dilute the bounces, is how senders blow through a 5% line in one campaign.

Step 4: keep catch-all sends contained

Even when the maths says yes, structure the send so a bad result hurts as little as possible.

1. Never send catch-all addresses first. Start a campaign with confirmed addresses so early bounce signals are clean.
2. Spread them across days, not one burst. A gradual rate gives you time to stop if bounces spike.
3. Suppress every hard bounce immediately. A catch-all address that bounced is now a confirmed hard bounce and must never be mailed again.
4. Keep the segment separate in reporting, so you can see its bounce rate on its own rather than averaged into everything else.
5. Watch complaints, not only bounces. A catch-all address that accepts mail may be a shared mailbox nobody asked to hear from you. Amazon SES's same page puts its complaint review line at 0.1%.

What not to do

Mistake	Why it hurts
Sending every catch-all address because "most are probably real"	You have no measurement behind "most", and one bad segment can push the whole account over its limit
Deleting every catch-all address	On B2B lists you throw away a large share of real people at companies that simply configure their mail this way
Re-verifying catch-all addresses hoping for a different answer	A catch-all domain will accept again. The result will not change until the company changes its configuration
Trusting a tool that reports catch-all addresses as valid	It is reporting the domain's setting, not the mailbox. See how to tell if your verifier is lying
Treating "unknown" the same as catch-all	Unknown is a failed check and may resolve on a retry. See verification results explained

How SimpleVerifier reports them

SimpleVerifier detects catch-all domains and returns those addresses as risky, never valid, and exports them separately from valid and invalid so the segment is ready to score. Because there are no per-address credits, re-checking a list after a company tightens its mail configuration does not add to the bill.

Practical takeaway

Catch-all is a sourcing question disguised as a verification question. Rank the addresses by where they came from, measure the best group with a test batch of at least 100, read the top of the confidence range, and use the formula to cap how many you send next to your confirmed addresses. To see how an individual address on your list comes back before you build the segments, use the free single email checker.