Article
Which Email Verifier Is the Most Accurate? How to Test It on Your Own List
By Unlimited Verifier Team ·

Summary
No single email verifier is the most accurate for every list, and vendors rarely publish comparable figures. The reliable answer comes from testing two or three tools on the same labelled sample of about 500 to 1,000 addresses you have already mailed. Judge them on false positives, dead-address catch rate and unknown rate together.No single email verifier is the most accurate for every list. The best one for you is the tool with the lowest false-positive rate on a labelled sample of roughly 1,000 addresses you have already mailed.
Vendors rarely publish comparable figures, and public documentation cannot rank the major tools. You can still get a defensible answer. Define accuracy in terms your campaigns feel, test two or three verifiers on the same labelled sample, and choose the one whose mistakes cost you least. Our pillar guide on email verifier accuracy covers the wider picture. This article focuses on the decision and how to make it.
Quick answer
- Measure two errors, not one. Track how often a Valid result bounces (false positives), how many dead addresses the tool catches, and how many addresses it leaves Unknown.
- Test on your own mailings. Use 500–1,000 addresses with known outcomes, and run every tool on the same file on the same day.
- Expect noise at small samples. With about 750 Valid-labelled addresses, a gap of around 1 point can be chance. Check the margin before you pick a winner.
- Treat vendor numbers as claims. Unlimited Verifier states 99.5% accuracy for deep verification. That is the company's figure, not an independent benchmark.
- Hold back uncertain addresses. Keep Risky and catch-all addresses off your main send until a test shows they perform.
What does "most accurate" actually mean for an email verifier?
Accuracy means a verifier's label matches what happens when you send. A single accuracy percentage hides two different failures, so measure them separately.
- A false positive is an address marked Valid that bounces. It damages sender reputation, which is why deliverability teams watch it most closely.
- A false negative is an address marked Invalid that actually receives mail. It costs you reachable contacts, which growth teams feel most.
A third measure is the unknown rate: the share of addresses the tool cannot resolve into Valid or Invalid. Unknowns are not errors, but each one still needs a decision. For the reasoning behind the first two measures, see false positives vs false negatives in email verification, and for the general accuracy picture, see how accurate email verification really is.
| Metric | Calculation | Question it answers | Who cares most |
|---|---|---|---|
| False-positive rate | Valid-labelled addresses that were undeliverable ÷ all Valid-labelled | "If it says Valid, will it bounce?" | Deliverability and compliance teams |
| Dead-address catch rate | Undeliverable addresses labelled Invalid ÷ all undeliverable | "How many dead addresses slipped through?" | Owners cleaning old databases |
| False-negative rate | Deliverable addresses labelled Invalid ÷ all Invalid-labelled | "Am I discarding reachable leads?" | Sales and growth teams |
| Unknown rate | Risky or unresolved ÷ total | "How much still needs a decision?" | Everyone |
A verifier can look strong on one score by being cautious and labelling many addresses Risky, or by being permissive and labelling most addresses Valid. Reading all four metrics together shows which trade-off you are buying.
Bottom line: Judge a verifier on its false-positive rate, dead-address catch rate and unknown rate together, never on one headline percentage.
Why can't one verifier win on every list?
Most verifiers rely on an SMTP mailbox check. The sender asks the receiving server whether it will accept a recipient during the RCPT TO step, as described in RFC 5321. An acceptance shows that the server accepted that address at that moment. It does not prove a person reads that inbox.
That gap explains most differences between tools:
| Address type | What the check can usually establish | Typical result | Why tools diverge |
|---|---|---|---|
| Mailbox at a domain that answers probes | Existence, usually | Valid or Invalid | Most tools agree here |
| Large consumer providers | Often limited; probes may be throttled or answered non-committally | Frequently Unknown | Retry logic and probe handling differ |
| Corporate catch-all domain | Server accepts any address | Risky | The specific mailbox cannot be confirmed |
| Disposable domain | Mailbox works but is throwaway | Flagged | Depends on each tool's disposable list |
| Role address (info@, sales@) | Mailbox exists; engagement usually low | Often Valid | Deliverable but weak; flag separately |
| Domain with no MX record | Domain cannot receive mail | Invalid | Easy; tools agree |
| Spam trap | Looks like a live mailbox | Often Valid | Generally not identifiable from SMTP alone |
The spam-trap row sets a ceiling on every verifier. Recycled traps are designed to look live, so a mailbox probe will often call them Valid. Catching them depends on reputation and complaint data. M3AAWG's best practices for senders cover list hygiene and trap-related damage, and no single SMTP check should be your only control.
Tools agree on clear cases and diverge on hard ones. An accuracy test that samples mostly easy addresses will flatter every vendor, so over-sample the hard segments.
Bottom line: Verifiers disagree most on address types with no clean answer, so their differences show up in your riskiest segments.
Why do vendor accuracy numbers differ, and how should you read them?
Vendor accuracy figures depend on choices that are rarely disclosed: which addresses were tested, what counted as truth, and whether Risky or Unknown results were scored as wrong or excluded. Change one choice and the same tool's score moves. The worked example below shows one tool's headline score ranging from 85.0% to 93.4% depending only on how Risky results are treated. For more on this problem, see how email verifiers measure accuracy.
Ask five questions of any accuracy claim:
- How big was the sample, and how was it drawn? A random sample from a live list differs greatly from a curated set.
- What was the ground truth? Bounce-confirmed delivery outcomes, hand-checked mailboxes, or the vendor's own check?
- How were Risky, catch-all and Unknown results scored?
- Which address types were included?
- When was it tested? Mailbox providers change behaviour over time.
Unlimited Verifier states 99.5% accuracy for deep verification. Treat that as a starting claim your own test should confirm, and ask the company for the sample size and ground truth behind it before relying on it. Are vendor accuracy claims for email verifiers trustworthy? goes deeper on this question.
If you need one number to compare tools, use the false-positive rate on a labelled sample from your own list. It is the number you can reproduce.
Bottom line: A vendor's accuracy percentage is only as good as its test design, so compare tools on your own data.
How do you test an email verifier's accuracy yourself?
Follow this framework. It takes an afternoon and produces a number you can defend. For a fuller version, see how to test an email verifier's accuracy yourself.
- Build the sample from addresses you actually mailed. Pull 500–1,000 addresses from campaigns in the last 6–12 months. Stratify them by type: consumer, corporate, suspected catch-all, and dormant (no engagement for more than 12 months). Include enough from the hard segments to test them properly.
- Define ground truth from delivery outcomes. Mark an address undeliverable if it produced a hard bounce, which is an SMTP enhanced status code beginning with 5, as defined in RFC 3463. Mark it deliverable if it was mailed and produced no hard bounce. Exclude addresses with repeated soft bounces or no send, because you cannot label them honestly.
- Run every tool on the identical file on the same day. Results shift as providers change behaviour and caches update.
- Build a confusion matrix. Put deliverable and undeliverable as rows, and Valid, Risky and Invalid as columns.
- Calculate the four metrics from the first table.
- Check the margin of error. For about 750 Valid-labelled addresses, a 95% margin on a 3% rate is roughly ±1.2 points. Two tools one or two points apart may be statistically indistinguishable, so enlarge the sample before choosing a winner on a small gap.
- Set a decision rule before you look at results. For example: "Valid must bounce at under 3%, and we must catch at least 70% of dead addresses." A rule set in advance stops you from picking the tool you already prefer.
Two caveats. Soft bounces, full mailboxes and provider blocks are not clean ground truth, so leave them out or label them separately. Email opens are weak ground truth too, because Apple Mail Privacy Protection and image blocking distort them.
Bottom line: A labelled sample from your own mailings, scored against a rule you set in advance, beats any published benchmark.
What does a worked accuracy test look like with real numbers?
This example is illustrative. Tools A, B and C are hypothetical, and the figures are invented to show the arithmetic. The sample is 1,000 addresses from one campaign: 780 were deliverable and 220 hard-bounced.
Confusion matrices
| Tool | Truth | Valid | Risky | Invalid |
|---|---|---|---|---|
| A | Deliverable (780) | 700 | 60 | 20 |
| A | Undeliverable (220) | 40 | 30 | 150 |
| B | Deliverable (780) | 740 | 30 | 10 |
| B | Undeliverable (220) | 20 | 40 | 160 |
| C | Deliverable (780) | 760 | 5 | 15 |
| C | Undeliverable (220) | 30 | 15 | 175 |
Metrics from the matrices
| Metric | Tool A | Tool B | Tool C |
|---|---|---|---|
| Labelled Valid | 740 | 760 | 790 |
| False-positive rate | 40 ÷ 740 = 5.4% | 20 ÷ 760 = 2.6% | 30 ÷ 790 = 3.8% |
| Dead-address catch rate | 150 ÷ 220 = 68.2% | 160 ÷ 220 = 72.7% | 175 ÷ 220 = 79.5% |
| False-negative rate | 20 ÷ 170 = 11.8% | 10 ÷ 170 = 5.9% | 15 ÷ 190 = 7.9% |
| Unknown (Risky) rate | 90 ÷ 1,000 = 9.0% | 70 ÷ 1,000 = 7.0% | 20 ÷ 1,000 = 2.0% |
| Headline accuracy, Risky counted wrong | 850 ÷ 1,000 = 85.0% | 900 ÷ 1,000 = 90.0% | 935 ÷ 1,000 = 93.5% |
| Headline accuracy, Risky excluded | 850 ÷ 910 = 93.4% | 900 ÷ 930 = 96.8% | 935 ÷ 980 = 95.4% |
What the numbers say
- Tool C has the highest headline score when Risky counts as wrong, yet it lets more bad Valid addresses through than Tool B (30 against 20). The headline hides that trade-off.
- Tool B has the lowest false-positive rate. Its gap to Tool C (2.6% vs 3.8%) is about 1.2 points, within the noise for this sample. Its gap to Tool A (5.4%) is clearly larger.
- If this sample reflects your list, 10,000 addresses labelled Valid would produce roughly 540 undeliverable with Tool A, 260 with Tool B and 380 with Tool C. Even Tool B would exceed a 2% bounce target, so Risky handling and re-engagement still matter. Does a valid result mean the email will be delivered? explains why a Valid label is a probability, not a guarantee.
Testing within Unlimited Verifier's limits
Running this 1,000-address sample on the Free plan takes seven lists (six of 150 and one of 100), within the 25-list monthly limit. Standard verification is unlimited on every plan, with no per-email fees. Deep verification is the constraint: Free includes 750 a month, so deep-verifying all 1,000 addresses would exceed the allowance by 250 this month.
On Pro, 1,000 deep verifications use 4% of the 25,000 monthly allowance. At full use, the email verification pricing works out to $29 ÷ 25 = $1.16 per 1,000 deep verifications on monthly billing, or $25 ÷ 25 = $1.00 on annual billing. Your effective cost rises if you use less than the full allowance, so spend deep verifications on the addresses you are least sure about.
Bottom line: The right tool depends on which error you can afford. Compare false-positive rate against your bounce tolerance, and catch rate against how many dead addresses you can tolerate.
Which verifiers belong in your test?
Choose tools by how their results map to your decisions, not by reputation alone. The table reflects public documentation as we read it. We have not tested these products, and status names change, so check current documentation before you build a mapping.
| Tool | How results are categorised (per public docs) | Differentiators | Watch-outs |
|---|---|---|---|
| Unlimited Verifier | Valid, Risky or Invalid; flags catch-all, disposable and role-based addresses | Open-source verification engine on GitHub; company-stated 99.5% deep-verification accuracy; API accepts batches of up to 10,000 emails per request | Free results are kept for 24 hours and Free has no API access; no native ESP, CRM or Zapier integrations, so the workflow is export, verify and re-import, or API calls |
| ZeroBounce | Valid and invalid, plus catch-all, unknown, spam-trap and abuse statuses | Dedicated spam-trap and abuse statuses, useful if traps or complaints are your main concern | More statuses to map before import; each one needs a documented rule for your sends |
| NeverBounce | Valid, invalid, disposable, catch-all and unknown | Separate disposable and catch-all classes, with unknown kept apart from invalid | The unknown class forces a decision on each run: hold, retest or send |
From public documentation alone we cannot rank these tools. The answer for your list is the output of the test above. As a rule of thumb, the tool with the lowest false-positive rate on your Valid segment is the safer choice for sending. The tool with the highest dead-address catch rate is the better choice for list hygiene. The difference between vendors often comes down to how they label the middle ground, which is why why ZeroBounce and NeverBounce disagree on the same email is worth reading before you map statuses. For a wider comparison, see our roundup of the best email verification tools and how to choose an email verifier.
Bottom line: Run your shortlist through the same labelled sample, and let the false-positive rate on your data decide the send list.
How should catch-all, Gmail and Outlook results change your decision?
Catch-all domains accept mail for any address, so the mailbox check cannot confirm a specific person. Treat a tool that calls catch-all addresses Valid with high confidence as suspicious, and test catch-all addresses as their own stratum. The catch-all email verification page explains how these results are handled, and how accurate email verifiers are on catch-all domains covers the testing in more depth.
Large consumer providers are the other common source of divergence. Results for Gmail and Outlook addresses can differ from one tool to the next, so read them separately. See email verifier accuracy on Gmail vs Outlook addresses for how to interpret them.
Our working rule is that a tool leaving more than about 10% of a consumer-heavy list Unknown is not ready for send decisions. That is a practitioner threshold, not an industry standard, so set your own before testing.
Verification also does not establish consent. For baseline obligations in the US, see the FTC's CAN-SPAM compliance guide. Email verification compliance and hygiene explains how verification fits into a wider process.
Bottom line: Test catch-all and consumer segments separately, because that is where tools diverge most and where a confident-looking Valid is least reliable.
What should you do with the results?
Use the test output to choose an action, not just a winner.
| Pattern in your results | What it means | Action |
|---|---|---|
| False-positive rate under 3% and catch rate above 75% | Strong on both measures | Use as primary verifier; hold Risky addresses back |
| False-positive rate above 5% | Valid labels are not safe for sending | Do not mail Valid addresses without a re-engagement step; tighten your rule or switch tools |
| Catch rate under 70% | Many dead addresses survive | Deep-verify the suspicious segment, or purge long-dormant addresses |
| Unknown rate above 10% | The tool is not resolving enough | Retest with deep verification and hold unresolved addresses; see what is a good unknown rate for an email verifier |
| Tools agree on about 95% or more | Differences sit at the margin | Focus on the disagreement set and choose on workflow fit |
These thresholds are our working rules, not published standards. The disagreement set is usually a small share of addresses, but that is where accuracy is decided, so review those addresses one by one before committing. Once you choose a tool, keep the list healthy with email list maintenance best practices, and consider real-time email validation at signup so bad addresses never enter the list. Teams running email verification for ecommerce and SaaS often find signup-form addresses are the main source of Risky results, so test those separately.
Bottom line: Let the pattern in your matrix set the action, and spend your attention on the addresses the tools disagree about.
How do you run the accuracy test in Unlimited Verifier?
- Build a labelled file. Export the addresses you mailed, and add a column called
truthwith the valuedeliverableorundeliverablefrom your campaign outcomes. Exclude addresses with repeated soft bounces. Save the file as CSV. - Plan your lists. Free allows up to 150 emails per list and 25 lists a month, so 1,000 addresses need seven lists. Pro allows up to 25,000 emails per list, so the same sample fits in one.
- Sign up for an account. The Free plan is enough for a small test.
- Upload and verify a list. Accepted files are CSV or Excel up to 10MB, and the email column is detected automatically.
- Run standard verification on the full set. It is unlimited on every plan. Then send only the Risky and catch-all addresses for deep verification, which draws on your 750 Free or 25,000 Pro allowance.
- Export results immediately as CSV or JSON. Free results are kept for 24 hours and Pro results for 30 days, so save your files before the window closes.
- Join results to your truth column by matching on email address, using XLOOKUP or VLOOKUP. If the result is in column C and the truth label is in column B on the same row, then
=COUNTIFS(C:C,"Valid",B:B,"undeliverable")gives your false-positive count. Divide it by=COUNTIF(C:C,"Valid"). - Repeat the test with other tools on the same file. For larger batches, the email verification API accepts up to 10,000 emails per request. API access is included in Pro, which covers 100,000 API emails a month, and is not available on Free. For background on automation, see email verification API and automation.
- Investigate disagreements individually with the free email checker, one address at a time.
Bottom line: The test runs in an afternoon, and its output is a false-positive rate you can take into any vendor conversation.
What common mistakes distort accuracy tests?
- Testing only on a clean list. Easy addresses flatter every tool. Over-sample catch-all, dormant and consumer segments.
- Quoting one headline percentage. Report the false-positive rate, catch rate and unknown rate together.
- Silently counting Risky as correct or wrong. Decide the treatment before you test, and state it in your report.
- Using opens as ground truth. Privacy features inflate and distort opens.
- Testing once. Results drift, so re-test quarterly if the list drives revenue.
- Running tools at different times. Run every tool on the same file on the same day.
- Ignoring spam traps. An SMTP check cannot see every trap, so pair verification with complaint and engagement monitoring.
- Losing results to retention windows. Export before the Free plan's 24-hour window closes.
- Treating Valid as consent. A deliverable address is not an opted-in address.
Bottom line: Many distorted tests come from a weak sample or a headline number chosen after the fact, not from the tools themselves.
How we compared (and limitations)
This article is based on public documentation, including RFC 5321 and RFC 3463, M3AAWG's sender guidance, and the FTC's CAN-SPAM guide. We have not run controlled benchmarks of any verifier against a live list, and we make no ranking claim. The three tools in the worked example are hypothetical, and their figures are illustrative. Vendor status categories change, so check each product's current documentation. Unlimited Verifier's 99.5% deep-verification figure is the company's own statement. For how verifiers define and report accuracy in general, see how accurate email verification really is.
Bottom line: Treat every vendor figure here as a claim to verify on your own list, not a ranking.
Five-step verifier accuracy test
- Pull 500-1,000 addresses you already mailed, with recorded outcomes (delivered, hard bounce, soft bounce).
- Run two or three verifiers on the same file on the same day.
- For each tool, calculate false-positive rate, dead-address catch rate, false-negative rate and unknown rate.
- Check the margin of error; treat gaps near one point as a tie.
- Choose the tool whose mistakes cost you least, weighting bounces against lost leads.
Frequently asked questions
Which email verifier is the most accurate overall?
No single email verifier is the most accurate for every list. Public documentation cannot rank the major tools because vendors rarely publish comparable figures. The most accurate option for you is the one with the lowest false-positive rate on a labelled sample of addresses you have already mailed.
What is a false positive in email verification?
A false positive is an address the verifier marks as Valid that then bounces when you send. It damages sender reputation, so deliverability teams track it most closely. Measure it as Valid-labelled addresses that were undeliverable divided by all Valid-labelled addresses.
What is a false negative in email verification?
A false negative is an address marked Invalid that actually receives mail. It causes you to discard reachable contacts. Measure it as deliverable addresses labelled Invalid divided by all Invalid-labelled addresses, which matters most to sales and growth teams.
How many addresses do I need to test email verifiers?
Most teams start with 500 to 1,000 addresses that have known outcomes, and run every tool on the same file on the same day. Small samples are noisy, so a gap of around one percentage point can be chance. Check the margin of error before declaring a winner.
Should I trust a vendor's accuracy percentage?
Treat vendor accuracy figures as claims rather than independent benchmarks. A company's stated accuracy often reflects its own test conditions. Run your own labelled test on your own mailing data to get a figure you can defend.
What should I do with Risky or catch-all addresses?
Keep Risky and catch-all addresses off your main send until a test shows they perform acceptably. Many teams send them to a separate, lower-volume segment first. This protects sender reputation while you learn how they behave.
What is an unknown rate in email verification?
The unknown rate is the share of addresses a verifier cannot resolve into Valid or Invalid, often because the mail server gives no clear answer. Unknowns are not errors, but each one still needs a decision about whether to send, hold or remove it.
How do I label addresses for an accuracy test?
Use addresses you have already mailed and record the actual outcome, such as delivered, hard bounced or soft bounced. Those outcomes become your ground truth. Avoid relabelling with a verifier, because that would test the tool against itself.
Does the best verifier change between lists?
Yes. Accuracy depends on list source, region, domain mix and age, so a verifier that performs well on one database may rank differently on another. This is why testing on your own data matters more than any published ranking.
Can I compare verifiers without running a full test?
You can check a free trial or small credit pack, but a meaningful comparison still needs a labelled sample with known outcomes. Without ground truth, you are only comparing labels, not accuracy.
Cut Bounces and Protect Your Sender Reputation With Unlimited Email Verification
Every invalid address you send to hurts your sender reputation and your inbox placement. Verifying your list before each send means fewer bounces, healthier deliverability, and a predictable cost. With Unlimited Verifier, standard verification is unlimited on every plan, with no per-email fees.
Unlimited Verifier checks syntax, domain, MX records and the SMTP mailbox. It also detects catch-all domains and flags disposable and role-based addresses. Export a CSV from your email platform or CRM, upload and verify the list, then re-import the clean results. You can also automate checks with the email verification API.
| Free | Pro | |
|---|---|---|
| Price | $0 | $29/month ($25/month billed annually) |
| Lists | 25 a month, up to 150 emails each | Unlimited, up to 25,000 emails each |
| Deep verifications | 750 a month | 25,000 a month |
| API access | No | Yes (100,000 emails a month) |
| Result retention | 24 hours | 30 days |
Why teams choose Unlimited Verifier
- Unlimited standard verification on all plans, including Free, with no per-email fees.
- Catch-all detection so you know which domains accept every address.
- Clear Valid / Risky / Invalid results, exportable as CSV, Excel or JSON.
- An open-source verification engine you can inspect on GitHub.
Unlimited Verifier states 99.5% accuracy for deep verification.
Create a free account and verify your next list before you hit send.
Want to test a single address first? Try the free email checker.