Email Deliverability
November 18, 2026

Is NeverBounce Accurate? We Tested It on 10,000 Catch-All Emails Against Allegrow

We tested 10,000 catch-all addresses. NeverBounce found 151 of 500 real contacts, Allegrow 491. See what its accept-all and unknown results mean.

Email Domain Sender Reputation Cover
Get a Free 14-Day Trial
Identify valid & invalid contacts on enterprise and catch-all servers with precision on up to 1,000 records.
Try Free Today

Table of Contents

NeverBounce says it provides up to 99.9% accuracy. So why does a corporate list come back from NeverBounce mostly labeled accept-all or unknown? More importantly, what should a sales team do with those records? We tested NeverBounce against Allegrow on 10,000 catch-all records to find out. NeverBounce rarely got an answer wrong. It also gave very few answers. It left most of the list unresolved, and still called more invented addresses valid than Allegrow did.

NeverBounce returned most of the list as accept_all_unverifiable, a label that confirms the domain accepts mail but says nothing about the person. To see what that label hides on your own data, take one address NeverBounce returned that way and run it through our free email verifier.

Below, we cover how we built the test, what NeverBounce’s results and statuses mean, how Allegrow performed on the same file, the cost per real contact found, and what to do with a large accept-all bucket. For NeverBounce pricing tiers, user reviews and other tools, see our NeverBounce alternatives guide.

‍

Key takeaways

  • What we tested: 10,000 addresses across 251 catch-all domains, made up of about 5,000 permutations for 500 real professionals plus 5,000 invented addresses, run through NeverBounce and Allegrow in August 2026.
  • Real contacts found: NeverBounce found 151 of 500 real people, fifth place among the eleven tools in our benchmark. Allegrow found 491.
  • Unresolved: NeverBounce left 96.3% of records unresolved, the highest rate of the 14 tests in the study. 8,947 of those 10,000 records came back accept_all_unverifiable.
  • False positives: declining to decide didn’t produce a low error count. NeverBounce called 27 of 5,000 invented addresses valid, against Allegrow’s 11. That means 27 of its 214 valid results were fictional people.
  • Cost per real contact: NeverBounce charges for every record, accept-all and unknown included, so cost per real contact found on this list was $0.32, against $0.28 for Allegrow.
  • What to do: on corporate lists, an accept_all_unverifiable result hands the decision back to the sender. Re-verify those records with a tool that is able to give a conclusive status rather than suppressing them or sending to them blind.

‍

How we tested NeverBounce on 10,000 catch-all email addresses

A verification tool can fail on a catch-all domain in two ways: (1) it can mark an address valid when no mailbox exists, or (2) it can decline to decide and hand the address back without an answer. Measuring both needs a list where the right answer is known in advance. So we built one.

The list holds 10,000 addresses across 251 domains, and every domain is catch-all, confirmed before testing began. They include Google, Stripe, Netflix, Databricks and Nvidia, the kind of corporate accounts B2B teams target every day. The records fall into two groups:

  • About 5,000 permutations for 500 real people. We took 500 professionals from public LinkedIn employee data at those domains and generated around ten standard formats for each one (firstname.lastname@, flastname@, f.lastname@, firstname@ and so on). At least one permutation per person should reach a real mailbox.
  • 5,000 invented addresses. These use obvious fictional character names at the same domains, so no mailbox should exist for any of them.

NeverBounce and Allegrow each ran on the identical file in August 2026. The same records went through 11 tools in 14 tests, and the full benchmark publishes every result. We scored three numbers:

Metric Denominator How it’s counted
Real contacts found 500 people A person counts as found if any one of their permutations comes back valid
False positives 5,000 invented addresses Any invented address marked valid
Unresolved 10,000 records Any result that isn’t a clear valid or invalid, whatever the vendor calls it

‍

Why test multiple permutations?

Sales teams and data providers rarely know a company’s exact email format. They generate a few likely versions of each address and rely on verification to pick the one that works. Testing permutations reproduces that job. The verifier has to find the live address among plausible alternatives, not just confirm one it was handed.

That’s also why real contacts found is counted per person. A rep doesn’t need all ten permutations of a prospect’s name to work. They need the one that does.

Why we only count false positives against the invented addresses

The permutation set doesn’t give a clean answer at the address level. We know each person is real, but not which of their ten candidates is the live mailbox, and aliases mean several can work at once. A valid result on a permutation therefore can’t be scored as a mistake.

The invented addresses are different. It’s highly unlikely that a person named after a famous fictional character works at these companies, so a valid result there is an error we can identify with confidence. That makes the invented set the only clean measure of false positives in the study, and every false-positive figure in this review is out of 5,000.

Why this methodology works for catch-all domains

Most verification comparisons use lists where nobody knows the true answer. Was an address marked “risky” correctly, or just cautiously? Did a bounce come from a dead mailbox or a temporary server problem? Without ground truth, a comparison can only rank tools by the share of results they label valid, and that isn’t the same as accuracy.

Planting known-real and known-invented addresses in the same list fixes that. Every result can be checked against the right answer, so the test measures what a team cares about: how many real people a tool confirms, and how often it confirms someone who doesn’t exist.

Limitations, stated plainly

Every domain in the test is catch-all by design, so the results describe catch-all performance specifically. A list of companies on standard mail setups would produce very different numbers for every tool, NeverBounce included.

Each tool ran once. Some results from both NeverBounce and Allegrow can resolve on a later run, because inconclusive outcomes often come from timeouts or temporary server responses. The real people were identified from public data rather than reply history. When you run this test yourself, contacts who have replied to you are a stronger source of ground truth, so use them. And it’s a point-in-time result: domain setups and verification products both change.

‍

Is NeverBounce accurate? Results on catch-all domains

Metric NeverBounce
Real contacts found 151 of 500 (30.2%)
False positives 27 of 5,000 (0.5%)
Unresolved 9,633 of 10,000 (96.3%)
Price for 10,000 verifications $49 a month (Growth plan)
Records charged on this list All 10,000, whatever the result

‍

NeverBounce found 151 of the 500 real people. That places it fifth of the eleven tools in the study and well ahead of most of the tools that left the bulk of the list unresolved. NeverBounce states that it provides “up to 99.9% accuracy”, and that claim and our result can both be true. Accuracy describes the answers a tool gave. Coverage (meaning, how many of the people you were looking for a tool actually found) describes how many it was willing to give.

On coverage, the gap is large. NeverBounce returned a clear valid or invalid result on 367 of the 10,000 records, fewer than four in every hundred, and left 96.3% unresolved. That’s the highest unresolved rate of the 14 tests in the benchmark.

Declining to decide on almost everything also didn’t produce a clean error rate. Other tools that left most of the list unresolved recorded very few false positives: Debounce 1 and EmailListVerify 2 to 3. NeverBounce called 27 invented addresses valid. Put another way, 27 of NeverBounce’s 214 valid results on this list were fictional characters, about one in eight. This means a valid label from NeverBounce on a catch-all domain may carry more risk than its headline accuracy figure suggests.

What do NeverBounce’s catch-all and unknown results mean?

NeverBounce returns five results. Its result codes guide marks two of them, accept-all and unknown, as “use caution”. On this list, those two made up almost the entire output:

NeverBounce result Records What NeverBounce says it means Counted as unresolved
accept_all_unverifiable (catchall in the API) 8,947 The domain accepts all mail, so the mailbox can’t be confirmed Yes
unknown 686 The server didn’t give a definitive answer in time Yes
valid 214 Verified as a real address No
invalid 153 Verified as not existing or not accepting mail No
disposable 0 A temporary, disposable address No

‍

Unknown is the smaller bucket and the more forgivable one. NeverBounce attributes it to greylisting or security gateways, and inconclusive results like these can often be settled on a later run. Allegrow’s some_risk status works the same way.

Accept_all_unverifiable is the weakness. It covered 89.5% of the whole list (8,947 of 10,000 records). It doesn’t mean the address is good or bad. It means NeverBounce stopped at the domain’s blanket “accept” response, which a catch-all server gives for a real employee and an invented name alike. In our test NeverBounce charged a credit for every record, so 89.5% of the credits spent on this list paid for a label that confirms the domain, not the person. NeverBounce’s own guidance for these addresses is to send selectively, in small batches, and to monitor bounces. That puts the verification decision back with the sender.

Why NeverBounce missed 349 of 500 real contacts

For 349 of the 500 real people, none of their roughly ten permutations came back valid. Most of their addresses sit in the accept_all_unverifiable bucket, with the rest unknown or marked invalid. A sales team working from NeverBounce’s output wouldn’t see these people as confirmed contacts at all.

As a rough illustration, at the rate we measured, a list of 100,000 real contacts at catch-all domains would leave about 69,800 people unconfirmed.

A missed real contact also never shows up as a problem. A false positive eventually appears as a bounce. A real prospect left in the accept-all bucket just gets suppressed or skipped, and nothing in a report shows that they were reachable. On corporate lists, the cost of low coverage is usually less visible than the cost of a high error rate, but it can be much larger.

‍

Allegrow results on the same catch-all list

Metric Allegrow
Real contacts found 491 of 500 (98.2%)
False positives 11 of 5,000 (0.22%)
Unresolved 197 of 10,000 (2.0%)
Price for 10,000 verifications $139 a month (Starter plan, with 5,000 additional contacts)
Records charged on this list All 10,000, whatever the result

‍

On the identical file, Allegrow found 491 of the 500 real people. That’s 340 more than NeverBounce and the most of any tool in the benchmark. It returned a clear valid or invalid result on 98% of the list, while calling 11 of 5,000 invented addresses valid, against NeverBounce’s 27.

Allegrow left 197 records unresolved, a 2.0% rate. This is a deliberate choice in Allegrow’s verification logic: when the mail server gives no reliable signal either way, Allegrow marks the record some_risk rather than defaulting it to valid or invalid. Those records are worth re-running after 30 days rather than sending to today. Allegrow bases every result on email server data rather than on defaults or third-party data sets, so a conclusive result reflects what the server showed about that mailbox.

That’s also where the two tools differ most. NeverBounce’s unresolved bucket is mostly the domain-level accept_all_unverifiable label. Allegrow resolves those catch-all addresses at the mailbox level. That leaves a small set of genuinely inconclusive records, which a later re-run can often settle.

‍

NeverBounce vs Allegrow: results and cost per real contact

On this list, Allegrow led NeverBounce on all three benchmark measures, and NeverBounce was cheaper per credit. Both tools charged for every record, so the fairest price comparison is what each one cost per real person it confirmed.

NeverBounce Allegrow
Real contacts found 151 of 500 (30.2%) 491 of 500 (98.2%)
False positives 27 of 5,000 (0.5%) 11 of 5,000 (0.22%)
Unresolved 9,633 of 10,000 (96.3%) 197 of 10,000 (2.0%)
How catch-all addresses are handled Labeled accept_all_unverifiable at the domain level Resolved to valid or invalid at the mailbox level in most cases
Price for 10,000 verifications $49 $139
Cost per real contact found on this list $0.32 $0.28
Best for Consumer and SMB list cleaning on a tight budget B2B teams verifying corporate and catch-all contacts for outbound

‍

NeverBounce costs about a third of Allegrow per credit: $0.0049 against $0.0139. But a credit spent on an accept_all_unverifiable label doesn’t produce a contact a team can email. Once you count only the people each tool confirmed, Allegrow cost less per real contact on this list, because it confirmed more than three times as many. For teams verifying well beyond 10,000 contacts a month, Allegrow also offers unlimited plans.

As a rough illustration, at the rates we measured, a list of 100,000 real contacts at catch-all domains would come back with about 30,200 confirmed by NeverBounce and about 98,200 by Allegrow. Coverage really starts to matter at that scale. The difference is whether most of a corporate list can be emailed with confidence, or whether it sits in a bucket somebody has to decide on by hand.

‍

What to do with NeverBounce catch-all and unknown results

If you’ve run a corporate list through NeverBounce, the accept_all_unverifiable and unknown bucket is probably the largest part of your results. On our list it was 96.3%. It’s large on corporate lists for a structural reason: in our census of all 500 Fortune 500 primary domains, 69% were catch-all, behind a secure email gateway, or both. The pattern is common across company domains, and those are the setups where a basic check stops at the domain.

A team holding that bucket has three options, and each one has a cost:

  1. Suppress it. Nothing in the bucket bounces, because nothing gets sent. On this list, though, suppressing it would have discarded 349 of the 500 real people, the prospects the list was built to reach.
  2. Send to it. The real people get your email, and so does everyone else in the bucket. On this list, at least 4,820 of the 5,000 invented addresses came back accept_all_unverifiable or unknown, so sending to the bucket means mailing thousands of addresses that don’t exist. The damage doesn’t always show up as bounces either. Gateway-fronted domains often accept the message and then bounce it later, or drop it without any bounce. Meanwhile the sending domain absorbs complaints and spam-trap risk that never appear in a bounce report.
  3. Re-verify it with a tool that resolves catch-all addresses at the mailbox level. This is the only option that separates the real people from the invented ones before anything is sent.

Re-verifying the bucket fixes one list. If the same problems come back with every new list, the cause is usually the verifier itself.

Watch for these signs: deals that stall because the right contact never got the email, bounce rates that climb whenever a campaign reaches corporate accounts, or replies drying up as more of your mail lands in spam. They usually trace back to the same cause: a verifier that labels catch-all addresses instead of resolving them.

At that point, switch to a verifier built for corporate domains. Allegrow’s catch-all email verification returns a conclusive valid or invalid result on catch-all and gateway-fronted addresses, so the contacts NeverBounce left as accept-all become prospects you can email or remove with confidence. For contacts already in outbound, Allegrow’s Safety Net works inside Outreach, Salesloft, Apollo and HubSpot, checking each one before a sequence step sends.

‍

Is NeverBounce right for B2B lists?

NeverBounce is accurate on the answers it gives. On a list built entirely from catch-all domains, though, it gave very few. It returned a clear result on fewer than four records in a hundred, found 151 of 500 real people, and still called 27 invented addresses valid.

For consumer and SMB lists, where most domains answer verification checks plainly, NeverBounce is a quick, low-cost cleaner. Corporate lists are different. There, the accept_all_unverifiable bucket is where most of the real prospects sit, so it needs resolving at the mailbox level, not suppressing or sending to blind.

The quickest way to see what’s in your own bucket is to test it. Take the records NeverBounce returned as accept_all_unverifiable or unknown, and run up to 1,000 of them through Allegrow on a 14-day free trial. You’ll see how many are real, reachable people that a legacy verifier left unconfirmed.

‍

Frequently asked questions about NeverBounce

What does “unknown” mean in NeverBounce?

It means NeverBounce couldn’t get a definitive answer from the mail server in time, often because of greylisting or a security gateway. The address is neither confirmed nor rejected. Unknown is different from accept_all_unverifiable, and a later re-run can often settle it. In our test, 686 of 10,000 records came back unknown.

What should you do with “accept_all_unverifiable” results in NeverBounce?

Don’t suppress them all, and don’t send to them blind. The label means the domain accepts every address, so the mailbox wasn’t confirmed. In our test it covered 89.5% of the list, including most of the real people. Re-verify those records with a tool that resolves catch-all addresses at the mailbox level.

How accurate does NeverBounce say it is?

NeverBounce says it provides up to 99.9% accuracy. That figure describes the answers it gives. On our 10,000-record catch-all list, NeverBounce returned a clear valid or invalid result on fewer than four records in a hundred, found 151 of 500 real people, and called 27 of 5,000 invented addresses valid.

Does NeverBounce charge for unknown and catch-all results?

Yes. In our test, NeverBounce used a credit on every record, whatever the result, including accept_all_unverifiable and unknown. That meant 89.5% of the credits spent on our list paid for a catch-all label that confirms the domain but not the person. Pricing at 10,000 verifications was $49 a month on its Growth plan.

Lucas Dezan
Lucas Dezan
Demand Gen Manager

As a demand generation manager at Allegrow, Lucas brings a fresh perspective to email deliverability challenges. His digital marketing background enables him to communicate complex technical concepts in accessible ways for B2B teams. Lucas focuses on educating businesses about crucial factors affecting inbox placement while maximizing campaign effectiveness.

Ready to optimize email outreach?

Book a free 15-minute audit with an email deliverability expert.
Book audit call