Deliverability insights
Page 3 of 15.
The Registrar-Email Exodus: GoDaddy Mail Halved in Four Years
GoDaddy-hosted mail fell 4.23% → 1.49% → 0.75% of top-1M domains. Why registrar bundles lose, plus a zero-loss migration checklist.
Should You Self-Host Email in 2026? 147,000 Top Domains Still Do
Self-hosted mail fell from 42.76% to 22.53% of top-1M domains, yet 147,712 still run their own MX. An honest checklist for when it makes sense.
Geopolitics in Your MX Record: Yandex 360's Rise and 70% Fall
Yandex-hosted mail peaked at 5.56% of top-1M domains in December 2019 and sits at 1.63% today. How macro events leave traces in public DNS.
The Commerce Email Stack: Shopify Now Sends for Nearly 1% of the Top-1M
Shopify appears in the SPF of 0.85% of top-1M domains, up from 0.04% in 2016. What the platform sends for merchants and what stays their job.
The State of DMARC 2026: 47% Enforced, 51% p=none — in 459,124 Records
DMARC across the Tranco top-1M in 2026: 459,124 records, 51% stuck at p=none, 47.16% enforced, and quarantine just overtook reject for the first time.
The 10-Lookup SPF Limit in the Wild: How Top-1M Domains Hit PermError
SPF allows 10 DNS lookups per check — then returns PermError. How the limit is counted, how multi-ESP stacks blow through it, and how to audit your record.
SPF Flattening Is Invisible to Every Survey — Including Ours (2026)
Flattened SPF records list raw IPs instead of vendor includes, so no DNS survey can attribute them to an ESP. Why every published ESP share is a floor.
A Field Guide to the Internet's Long Tail: 36,377 Unmatched MX Hostnames
Beyond Google and Microsoft: 36,377 MX hostnames our 310+ patterns cannot classify, covering 81,514 top-1M domains. A taxonomy of the unclassifiable.
+all and Other SPF Suicide Notes Found in the Top Million Domains
The qualifier on your final all mechanism is the whole policy. What +all really authorizes, why ?all is a shrug, and how to choose between ~all and -all.
DKIM Selector Archaeology: What selector1, k1 and mte1 Reveal
DKIM selectors are vendor fingerprints: selector1 means Microsoft 365, k1 means Mailchimp. How selector scans map sending stacks — and what they miss.
Null MX Done Right (and Wrong): How Top-1M Domains Say 'We Don't Take Mail'
RFC 7505 Null MX versus the hacks: 4,925 empty MX targets, 502 localhost records, 157 tildes in the top-1M — and what each pattern costs senders.
Email Is IPv4's Last Fortress: AAAA on MX Across a Million Domains
Only 4.96% of self-hosted MX hosts in the Tranco top-1M publish AAAA records, while Google's MX fleet is 90% IPv6 — why email remains IPv4's last holdout.
234,559 Domains Publish DMARC and Enforce Nothing: The p=none Illusion
234,559 of 459,124 top-1M DMARC records — 51% — sit at p=none, monitoring without enforcement. What attackers read into it and the staged path to reject.
Proofpoint and Mimecast Are Eating the Enterprise Email Perimeter
Proofpoint's MX footprint roughly doubled since 2022 (0.86%→1.91%) and Mimecast grew to 1.53%. What gateway consolidation means for admins and senders.
Your Apex Is Protected, Your Subdomains Are Not: The sp= Gap
A p=reject apex can hide an sp=none tag that exempts every subdomain from DMARC enforcement. How the gap appears, what attackers do with it, how to close it.
KnowBe4 in Your SPF: Phishing-Training Adoption as a Security-Maturity Signal
KnowBe4 appears in 0.56% of top-1M SPF records, 2.8× more than in 2022. Why phishing-training adoption is readable in DNS and what the signal really means.
DMARC Report Addresses Leak Your Security Vendor: The rua= Market
The rua= address in a public DMARC record names your monitoring vendor — Valimail, dmarcian, EasyDMARC and peers. What it discloses and how to audit the loop.
A Decade of Email Consolidation, Charted: 2016–2026 in 192 DNS Snapshots
Ten years of DNS data: self-hosted email halved to 22.53%, Google and Microsoft reached 38.4%, and Amazon SES became the #1 ESP.
What Happens When Two Companies Hold 38% of the World's Business Email?
Google and Microsoft now receive mail for 38.4% of top domains. Filters as private regulation, outage blast radius, and the case for calm.
The Weirdest MX Records in the Top Million Domains
MX records pointing at localhost, a literal ~, and nothing at all — plus a 70-character hostname. Every DNS curiosity comes with a lesson.
How We Classify a Million Domains' Email Stacks Every Night — and Where It Breaks
Our open methodology: OpenINTEL snapshots, 310+ MX patterns, 81.5% SPF coverage — and honest blind spots like flattening and dictionary jumps.
Brazil's Email Market Is 43% One Company: Regional Monopolies Hiding in Global Averages
Locaweb holds 42.58% of Brazilian ESP-classified domains and KingHost is 99.8% domestic. Country-level DNS cuts reveal what global charts hide.
Why Warmup Vendor Scores Don't Match Real Inbox Placement (2026 Data)
Fifty anonymised 2026 case files where vendor-reported warmup scores of 90+ coexisted with independent inbox placement of 40-60%. The measurement gap, why it exists, and how to close it.
The Conflict of Interest in Email Warmup
When the vendor selling the warmup service is also the party grading it, you do not have verification — you have marketing. A 2026 framing of the structural problem and what neutral verification looks like.
What Happened to Warmup Networks in 2024 (And Why It Still Matters in 2026)
A factual review of the 2024 Google and Yahoo Bulk Sender Rules, the public enforcement actions against warmup pools that followed, and how the 2026 placement landscape was shaped by both.
How to Independently Audit Your Email Warmup (2026 Methodology)
A reproducible audit methodology for any warmup tool: setup, controls, the eight-step measurement protocol, the report template, and how to defend the result to a vendor that disputes it.
Stop Trusting Your Warmup Score. Verify It.
A 2026 buyer-side stance: the vendor score is one input, not the truth. Run an independent verification on every campaign that matters. The reasoning, the cost, and the routine that makes verification a habit.
Mail.ru Spam Filter in 2026: A Practical Reverse-Engineering Guide
How Mail.ru actually decides Inbox vs Spam in 2026 — DKIM-alignment as a hard gate, domain-over-IP reputation, list-unsubscribe as a strong positive signal, and the differences from Gmail/Outlook that catch foreign senders by surprise.
SPF vs DKIM vs DMARC at 9 Mailbox Providers — Who Cares Most About What?
How Gmail, Outlook, Yahoo, iCloud, AOL, GMX, T-Online, Mail.ru and Yandex actually weigh SPF, DKIM and DMARC differently — and the per-provider config that minimises Spam-folder placement.
Inbox Placement at GMX and T-Online: The DACH Deliverability Guide Nobody Wrote
GMX and T-Online together hold a meaningful share of the German consumer inbox. Their filters are stricter on SPF, friendlier to CSA-whitelisted senders, and unforgiving of foreign DKIM domains. What to actually do.
Yandex Pochta vs Mail.ru: Side-by-Side Spam Filter Behavior
Yandex and Mail.ru together hold the CIS consumer inbox. Their filters look similar from the outside and behave very differently in practice — DKIM weighting, ARC handling, FBL signals, and per-recipient learning rates.
BIMI Across 9 Mailbox Providers: Who Actually Renders Your Logo (Spoiler: Not Mail.ru)
BIMI is sold as universal sender branding. In 2026 only four major providers actually render the logo, three are exploring it, and the CIS giants Mail.ru and Yandex are nowhere on the roadmap. Here is the real provider matrix.
Email Throttling at Mail.ru and Yandex: Why CIS Senders Hit Quotas Earlier Than at Gmail
Mail.ru and Yandex enforce per-IP and per-domain rate limits that bite earlier than Gmail or Outlook. The 4xx temp-fail patterns, the retry windows that work, and the volume ramp that does not get throttled.
iCloud Mail's Hidden Spam Folder: Why You Don't Even Know You're Filtered
iCloud Mail filters silently. There is no Postmaster Tools, no SNDS, and the Junk folder is buried two taps deep on iOS. The result is opaque deliverability. Here is how to actually see what Apple is doing.
Outlook Focused vs Other vs Junk: What Each Folder Actually Means for Marketers
Microsoft splits the inbox three ways: Focused, Other and Junk. The marketing implications are bigger than they look — Other is not failure and Junk is not always content. Here is the per-folder playbook.
Gmail Promotions vs Spam: Which Tab Hurts Your Open Rate More?
Promotions is not Spam. But Promotions can hurt open rate more than Spam in some segments because users actively visit Spam to recover legitimate mail and rarely open Promotions at all. The data and the playbook.
How Often Should You Test Inbox Placement? A Per-Sender Cadence Guide
Daily, weekly, monthly, or only on-demand? The right testing cadence depends on volume, audience, and rate of change. Here is the framework — plus the events that should always trigger an out-of-band test.
Course Platform Emails: Teachable, Thinkific, GetCourse Deliverability
Course-platform emails are sent from the platform's IP pool. Inbox rates vary widely. The custom-domain (CNAME) options at Teachable, Thinkific, and GetCourse — and what to test.
Zoom/Google Meet Invite Emails: Do Calendar Invites Bypass Spam?
Calendar invites (.ics) get auto-added to many users' calendars regardless of spam status. Why that's the most underused 'always delivers' channel for events.
Event Ticket Emails Not Received: Eventbrite, Luma, Timepad Deliverability
Eventbrite, Luma, Timepad all send tickets from their own domains. Eventbrite inboxes around 92%, Luma higher, Timepad weaker — what to do as an organizer.