Across 7,930,038 lead addresses verified between 13 April and 10 October 2026, business lists were 14.5% invalid and freemail lists 17.6% invalid. The spread between providers is the real story: gmail.com addresses were 3.5% invalid, Australian hotmail.com.au addresses 92.7%. If you buy, scrape or export leads, this page tells you what to expect before you send.
The corpus is every address submitted to BounceZero bulk verification by customers in the 180 days to 10 October 2026 that reached a final classification: 7,930,038 rows across 6,514,884 business addresses and 1,415,154 freemail addresses. Each address went through the full pipeline described on the product facts page: syntax, DNS and MX, SMTP and provider-specific checks, domain intelligence and scoring.
Four outcomes are reported here. Invalid means the mailbox was rejected or the domain cannot receive mail. Catch-all means the domain accepts any address, so the mailbox cannot be confirmed. Unknown means the provider blocked or did not answer the probe; BounceZero refunds these and does not count them as deliverable. Valid means the mailbox was confirmed. The four do not sum to 100 because risky, role and disposable addresses are reported separately.
Two customers' ISP-heavy lists make up over 3 million rows (t-online.de and alice.it). Provider rows are shown so you can discount them; the business and freemail totals include everything.
A 41% unknown rate on t-online.de says the provider throttles verification, not that the list is bad. Treat unknown as unproven, never as invalid.
A TLD is shown only with 5,000 or more addresses, a provider only with 20,000 or more. Smaller cuts are not published.
Freemail addresses on a B2B lead list are a known weak point, and the corpus confirms it: 17.6% invalid against 14.5% for business domains. The surprise is the unknown column. Business domains return unknown 27% of the time because corporate gateways and ISPs block probes; freemail providers answer almost always. A business list that reads "74% valid" from a verifier that hides unknowns is not comparable with one that reports them.
| List type | Addresses | Invalid | Catch-all | Unknown | Valid |
|---|---|---|---|---|---|
| Business domains only (consumer ISP and freemail excluded) | 1,269,919 | 22.9% | 9.4% | 22.5% | 40.3% |
| Business domains | 6,514,884 | 14.5% | 2.3% | 27.0% | 32.6% |
| Freemail domains | 1,415,154 | 17.6% | 1.1% | 2.1% | 74.0% |
Freemail = gmail, googlemail, yahoo, ymail, hotmail, outlook, live, msn, icloud, me, aol, proton, gmx, web, mail, yandex, zoho. Window: 13 April to 10 October 2026.
Country TLDs are a rough proxy for where a list was sourced. Australian lists stand out: 36.5% invalid and the highest catch-all share after .org, consistent with the churn of Australian ISP mailboxes (tpg.com.au, hotmail.com.au) that appear in the provider table below. German .de and Belgian .be lists are dominated by providers that block probes, so their unknown rates, not their invalid rates, are the number to plan around.
| TLD | Addresses | Invalid | Catch-all | Unknown | Valid |
|---|---|---|---|---|---|
| .au | 560,378 | 36.5% | 6.3% | 17.6% | 37.3% |
| .nz | 36,773 | 28.9% | 0.1% | 3.6% | 11.6% |
| .org | 56,401 | 24.6% | 12.4% | 25.0% | 36.2% |
| .nl | 87,309 | 21.9% | 0.2% | 4.5% | 54.6% |
| .net | 1,252,193 | 17.7% | 0.6% | 28.8% | 19.9% |
| .com | 2,158,955 | 16.4% | 4.4% | 14.7% | 59.7% |
| .de | 2,095,440 | 13.9% | 0.0% | 41.2% | 16.5% |
| .fr | 119,582 | 12.1% | 0.5% | 9.3% | 47.6% |
| .ch | 30,522 | 9.3% | 0.3% | 36.0% | 19.6% |
| .fi | 100,516 | 4.6% | 0.2% | 22.1% | 59.6% |
| .be | 28,040 | 4.6% | 0.5% | 49.5% | 40.5% |
| .it | 1,193,715 | 1.3% | 0.0% | 3.2% | 64.2% |
Sorted by invalid rate. Only TLDs with 5,000 or more addresses. .de, .it and .net are dominated by single ISP providers; see the provider table.
Providers decide most of a list's fate. Gmail addresses are almost always alive, because Google deletes little and verification works. Legacy ISP mailboxes are the opposite: att.net, sbcglobal.net and bellsouth.net sit on the same AT&T platform that blocks probes, so three to four in ten come back unknown. Australian hotmail.com.au is a graveyard, 92.7% invalid, which is what happens when a provider retires a domain and the lists keep circulating.
| Provider | Addresses | Invalid | Catch-all | Unknown | Valid |
|---|---|---|---|---|---|
| hotmail.com.au | 110,452 | 92.7% | 0.0% | 3.1% | 2.3% |
| telefonica.net | 159,794 | 42.8% | 0.0% | 8.0% | 48.5% |
| tpg.com.au | 246,811 | 30.1% | 8.9% | 5.1% | 53.9% |
| hotmail.com | 253,445 | 16.5% | 0.1% | 0.0% | 70.5% |
| t-online.de | 2,088,148 | 13.9% | 0.0% | 41.2% | 16.4% |
| att.net | 421,292 | 10.1% | 0.0% | 26.4% | 5.8% |
| sbcglobal.net | 232,302 | 4.6% | 0.1% | 36.3% | 18.0% |
| bellsouth.net | 134,384 | 4.6% | 0.1% | 43.7% | 15.5% |
| gmail.com | 801,225 | 3.5% | 0.0% | 0.0% | 96.0% |
| alice.it | 1,178,297 | 1.0% | 0.0% | 2.7% | 64.7% |
Providers with 20,000 or more addresses. Valid, invalid, catch-all and unknown do not sum to 100; the remainder is risky, role and disposable.
A B2B list that has not been verified within 90 days carries about one invalid address in seven. At 2% bounce rate thresholds on Google and Microsoft, that is enough to get a sending domain throttled in the first campaign.
Ask any seller for 200 to 500 random rows and verify them. Sellers send their best sample; if the sample is above the benchmark for its list type, the full list is worse.
A verifier that reports 'valid' on t-online.de or att.net addresses it never confirmed is guessing. Ask what share came back unknown and whether you were charged for it. BounceZero refunds every unknown.
Route ISP and catch-all addresses to warm, low-volume sequences and keep cold volume on confirmed mailboxes. The provider column above is the segmentation key.
.au lists at 36.5% invalid and hotmail.com.au at 92.7% should be re-verified on purchase, not on upload, and priced accordingly.
Addresses decay. The list decay study measured the monthly loss; a quarterly re-check keeps a list inside the bounce thresholds.
In our 2026 corpus, 14.5% of business addresses and 17.6% of freemail addresses were invalid at verification time. Lists from Australian ISP domains and from telefonica.net were far worse, 30% to 93% invalid. A bought list that has not been verified in the last 90 days should be assumed to carry at least 15% invalid addresses.
A catch-all address belongs to a domain that accepts mail for any local part, so the mailbox cannot be confirmed by SMTP. In this corpus 2.3% of business addresses were catch-all, rising to 12.4% on .org domains and 6.3% on .au. Keep them for warm outreach, keep them out of cold campaigns, and never count them as verified.
Unknown means the receiving provider blocked or did not answer the verification probe, so the address is neither proven valid nor invalid. It is a property of the provider, not the list: t-online.de and bellsouth.net return unknown for roughly four in ten addresses. BounceZero refunds unknown results and does not count them as deliverable.
Among the TLDs with at least 5,000 addresses in the corpus, Australian .au domains had the highest invalid rate at 36.5%, followed by .nz at 28.9%, .org at 24.6% and .nl at 21.9%. Italian .it domains had the lowest at 1.3%, driven by alice.it.
Ask the seller for a random sample of 200 to 500 addresses and verify it. If the invalid share is above the benchmark for that list type, the full list will be the same or worse, because sellers send their best sample. BounceZero gives every account 100 free checks a month, enough to test a sample.
Segment definitions: freemail is the 17 largest consumer mailbox brands; non-freemail is everything else and therefore includes consumer ISP mailboxes (T-Online, AT&T, Telefónica and others), which is why its rates sit between the two; business domains only additionally excludes about 170 known ISP and consumer mailbox domains, so it is the row to use for B2B planning. All rates are BounceZero probe outcomes on customer-uploaded lists over the window, not a census of mailboxes; unknown measures what verification could not confirm, which depends on the provider's gateway policy during the window.
Cite as: BounceZero, Lead List Quality Benchmarks 2026, bouncezero.io/lead-list-quality-benchmarks-2026, published 10 October 2026. Data refreshed with each new window; the as-of date above is authoritative.
Per-provider and per-country pages, each with month-by-month rates and handling advice: Lead list quality by provider and country.
Upload a sample or a full list. Every account gets 100 free checks a month, unknowns are refunded, and the result splits valid, invalid, catch-all and unknown exactly as reported here.
Ayoub built BounceZero's 5-stage validation pipeline, its dedicated BGP-announced IP infrastructure, and the Patroni HA PostgreSQL cluster behind every verification. Previously built high-volume email delivery infrastructure. Trained at 1337 Benguerir (École 42 network, 2019). Open-source: bgp_analyzer.
Where leads come from, how finders work, and what a verified lead is - with 2026 corpus data
How finders work, Hunter vs Apollo vs Snov vs Clay, finder vs verifier
Extensions, Sales Navigator exports, accuracy, the legal line
Verify an email list in bulk
Email finder for companies
Free email lookup in 2026
Clay data enrichment explained
Continue through related topics