The Inbox Infrastructure Report 2026 | BounceZero Research

BounceZero Research · Flagship Report

The Inbox Infrastructure Report 2026

What 272,446 domains reveal about the state of email authentication. Snapshot, August 2026 — every claim below is independently verifiable with dig.

By Ayoub Lebda, Founder ·August 2026 ·Data: 2026-02-15 → 2026-08-05
Sample frame, stated plainly: these are domains observed on commercial mailing lists processed by BounceZero’s verification platform — not a random sample of the DNS. The sample over-represents B2B and commercially-mailed domains. That is exactly why the numbers matter to senders: this is the population you are actually emailing. Statistics are aggregate-only; no individual domain or customer list is identified, and provider cells below n=1,000 are suppressed.

The five numbers

62.7%
publish any DMARC record
13.4%
actually enforce it (p=reject)
20.0%
of probed domains are catch-all
1.28%
have adopted MTA-STS
1.9×
Google Workspace vs M365 catch-all rate

Two-thirds of the domains businesses actually email are either unprotected or only watching.

Part 1 — The DMARC enforcement gap

Publishing DMARC is now mainstream. Enforcing it is not.

DMARC policy Domains Share
No DMARC record at all 102,402 37.6%
p=none (monitor only) 81,094 29.8%
p=quarantine 52,469 19.3%
p=reject (enforcing) 36,440 13.4%

p=none is where domains go to die. Nearly a third of all domains have done the hard part — published a record, pointed a DNS entry — and then stopped before the one step that actually stops spoofing. Contrast with SPF at 83.4% adoption: the industry embraced the easy control and stalled on the one with operational risk. The full funnel reads 83.4% SPF → 62.7% DMARC → 13.4% enforcement.

Footnote on real-world DNS hygiene: 41 domains publish a DMARC policy value so malformed it fits no known syntax — including several misspellings of “quarantine”.

Part 2 — The provider divide

Your mail provider predicts your authentication posture more than your size does.

Provider Domains Has DMARC Enforcing (p=reject) MTA-STS
Microsoft 365 100,536 71.0% 16.3% 1.53%
Google Workspace 61,456 69.2% 12.2% 1.51%
Gmail (consumer) 3,504 88.6% 21.9% 2.00%
Zoho 2,472 68.2% 8.9% 0.53%
IONOS 1,147 88.8% 3.2% 0.09%
Fastmail 1,078 53.2% 7.5% 2.60%

Microsoft 365 and Google Workspace are statistically twins on adoption (~70%), but M365 domains enforce at a rate a third higher (16.3% vs 12.2%). The sharpest anomaly is IONOS: 88.8% publish DMARC, only 3.2% enforce — the signature of a registrar auto-provisioning records that nobody ever tunes.

36.3% of observed domains resolve to no identifiable provider; that bucket is disclosed here and excluded from the table rather than renormalised away. Providers below n=1,000 are suppressed.

Part 3 — The catch-all problem

One in five probed domains accepts mail for addresses that do not exist.

Of 239,431 domains where a direct SMTP catch-all probe completed, 47,909 (20.0%) accepted mail for a deliberately nonexistent address. These are per-domain, direct-probe measurements — not inferences.

Provider Domains probed Catch-all rate
Gmail (consumer) 1,344 37.8%
Google Workspace 57,614 33.2%
Microsoft 365 40,598 17.6%
Zoho 1,330 7.2%

The finding worth quoting: Google Workspace domains are ~1.9× more likely to be catch-all than Microsoft 365 domains (33.2% vs 17.6%), on large and comparable samples of 57,614 and 40,598 probed domains. The operational consequence is direct — a list heavy in Google Workspace domains will return far more addresses that no SMTP probe can resolve, and any vendor claiming a clean binary valid/invalid verdict on those addresses is guessing. That includes us: this class of address cannot be resolved by SMTP probing, by us or by anyone.

Definition: a domain counts as catch-all when a direct SMTP probe for a deliberately nonexistent local part was accepted. Rates here are domain-weighted. Volume-weighted (per-email) rates differ substantially because catch-all domains skew small — our companion study The Catch-All Mirage covers that distinction in depth.

Part 4 — Authentication is a function of age

New domains are the least protected — the inverse of the legacy-infrastructure assumption.

Domain age Domains Has DMARC Enforcing
Under 1 year 4,211 40.5% 7.3%
1–5 years 26,661 45.3% 5.5%
5–10 years 35,767 61.0% 10.9%
10+ years 142,072 72.6% 16.8%

DMARC adoption climbs monotonically with age, from 40.5% on domains under a year old to 72.6% at ten-plus years. Newly-registered domains are disproportionately used for cold outreach — which means the domains most likely to be mailing you are the least likely to be protected.

Registration age is known for 208,711 of 272,446 domains (76.6% coverage); rows with unknown age are excluded from this table only.

Part 5 — The transport-security frontier

The modern standards have not arrived.

1.28%
MTA-STS adoption

3,488 of 272,446 domains publish an MTA-STS policy. Eight years after RFC 8461, strict transport security for email remains a rounding error.

35.3%
Valid reverse DNS

Barely a third of domains resolve to mail infrastructure with valid forward-confirmed reverse DNS — still a top silent killer of deliverability.

Part 6 — The junk supply

BounceZero currently tracks 195,199 active disposable and throwaway email domains, aggregated from three independent sources. We deliberately do not publish an “X% of your list is disposable” figure — that number is entirely a function of whose list happens to be processed, and any vendor quoting one without that caveat is selling, not measuring.

Methodology

Collection window: 2026-02-15 to 2026-08-05 (about 5.6 months). Every figure is a point-in-time snapshot as of August 2026, not an annual trend. DNS records are re-checked on a rolling basis.

Sample frame: 272,446 unique domains observed on commercial mailing lists processed by the BounceZero verification platform. This over-represents B2B and commercially-mailed domains and is not a random sample of the DNS — it is a census of the domains senders actually email.

Denominators: authentication percentages use all 272,446 domains. Catch-all percentages use the 239,431 domains where a direct SMTP probe completed with a determinate result. Age percentages use the 208,711 domains with known registration age.

Catch-all definition: a domain is catch-all when a direct SMTP probe addressed to a deliberately nonexistent local part was accepted by the receiving server. No inference, no pattern-matching.

Provider attribution: derived from MX records. 36.3% of domains have no identifiable provider; that bucket is disclosed and excluded from provider tables rather than renormalised. Provider cells below n=1,000 are suppressed throughout.

DMARC normalisation: policy values are lower-cased, trimmed, and bucketed; 41 unparseable values are counted as malformed and excluded from policy shares.

Verify it yourself: every DNS claim in this report can be reproduced for any domain with dig TXT _dmarc.<domain>, dig TXT <domain> (SPF), and dig TXT _mta-sts.<domain> — or with our free DMARC, SPF, and DKIM checkers.

Citation: data and figures are free to cite with attribution to BounceZero and a link to this page (CC BY 4.0).

Where does your list stand?

A fifth of the domains on a typical B2B list accept mail for addresses that do not exist. Verify yours before your next campaign — 100 free checks a month, no card required.

Verify your list free

Or check a single domain with the free email verifier.

Email verification & deliverability explained

Deep-dive guides on how email verification and inbox placement work