Vendors in this market publish high numbers. ZeroBounce claims 99.6%, Clearout 99.73% and Reoon over 99%. Bouncer and Emailable publish no accuracy percentage at all, as of September 2026.
Then there's the one public benchmark with its method stated. Hunter ran 3,000 real business addresses through 15 verifiers and published the results on 3 November 2025. Overall accuracy ran from 31.2% to 70.0%, and the top ten sat between 57.6% and 70.0%.
Both sets of numbers can be true because they divide by different things. A vendor figure usually covers only the addresses the verifier chose to judge. A benchmark that counts the whole list keeps the catch-all domains and the servers that refused to answer, which is where the hard work lives. Our post on the 99% accuracy problem shows the arithmetic.
So an accuracy claim isn't a fact you accept or reject. It's a statement you can audit, and the questions below turn it into one. More on how verifiers reach a verdict sits on the ZapBounce blog.
Five questions that audit any accuracy figure
First, what's the denominator? Ask whether the figure covers every address submitted or only those that got a definite verdict. If unknown and catch-all results were left out, the figure describes a smaller, easier list than yours.
Second, what was the coverage? Coverage is the share of the list that came back valid or invalid. A verifier that's right 99% of the time on 65% of a list has told you less than one that's right 95% of the time on 90% of it.
Third, how often did a valid result bounce? That's the false-positive rate, and it's the number that reaches your sender reputation. Ask for it on its own line.
Fourth, whose data was it? Ask for the size of the test list, whether it was business or consumer mail, when it was run and who picked the addresses. Hunter's own tool came top of Hunter's benchmark, which is worth knowing when you read any vendor's test, including the ones that publish their method.
Fifth, can you reproduce it? Build a sample from your own data: addresses that engaged in the last month, addresses that hard-bounced recently and a slice of catch-all domains. Run it through each verifier you're weighing and compare the results against what you already know. Then read the counts on what accuracy is actually achievable.
What a fair number looks like
The short version
- Ask what an accuracy figure divides by before you compare two vendors on it.
- Ask for coverage and the false-positive rate as separate numbers.
- Test each verifier on a sample of your own addresses whose status you already know.
Questions people ask
Are 99% email verification accuracy claims true?
They are usually true for the addresses the verifier chose to judge. Hunter's benchmark of 15 verifiers on 3,000 business addresses, published 3 November 2025, measured 31.2% to 70.0% once the whole list was counted.
How do I test an email verifier's accuracy myself?
Build a sample of addresses whose status you know: recent engagers, recent hard bounces and some catch-all domains. Run it through each verifier and compare their verdicts with what you already know about each address.
What is a good accuracy figure for an email verifier?
95% to 98% on domains that respond honestly is a defensible ceiling. Read it next to coverage, because a high accuracy figure on a small share of the list tells you little.
Which verifiers publish no accuracy percentage?
Bouncer and Emailable, as of September 2026. ZeroBounce claims 99.6%, Clearout 99.73% and Reoon over 99%.