orbis://guides/cold-email-deliverability
--:--:--
$ orbis --guide deliverability
[boot]read=9min · scope=cold-outboundok
#
Cold email deliverability: a practical guide
>
Deliverability is not a setting you switch on. It is the running score mailbox providers keep on your domain, and every send either adds to it or spends it. This guide covers the controls that actually move that score, in the order they matter.
SEC.01
How filtering works
>
Providers ask three questions about every message: can I prove who sent this, does this sender have a history of people wanting their mail, and does this specific message look like the mail people want. Authentication answers the first, reputation answers the second, content answers the third. Fail any one badly and you land in spam regardless of the other two.
signals.md
- identity
- SPF, DKIM, DMARC alignment
- reputation
- domain and IP history, complaint rate, bounce rate
- engagement
- opens, replies, moves out of spam, deletions
- content
- links, images, spoofed formatting, attachments
SEC.02
Authentication
>
Authentication is the cheapest deliverability work you will ever do, and it is binary — either the records are right or your mail is suspect from the first send.
-SPF: publish one record listing every service allowed to send for the domain
-DKIM: sign with a key your sending platform controls, and confirm the signature validates
-DMARC: publish a policy so providers know what to do with unaligned mail, and collect the reports
-Custom tracking domain: a shared tracking host inherits everyone else's reputation
-Reverse DNS and a valid From address that can actually receive replies
SEC.03
Domains and warmup
>
Never run cold outbound from your primary corporate domain. Register separate sending domains, point them at the brand, and treat them as consumables. A new domain has no reputation at all, so it must earn one before volume arrives.
01
Separate domains
Cold volume lives on lookalike domains so a reputation hit never touches your main mail.
02
Age before use
Let a new domain sit with authentication in place before it sends anything cold.
03
Ramp slowly
Start at a handful of sends per mailbox per day and increase gradually over weeks, not days.
04
Keep warming
Warmup is continuous background traffic, not a one-off phase you finish.
SEC.04
Volume and pacing
>
Scale by adding mailboxes, not by pushing more through each one. A mailbox has a natural ceiling; crossing it is the single most common way teams burn a domain in week three. Spread sends across the working day rather than firing a batch at 09:00, and keep per-mailbox limits enforced by the system rather than by discipline.
the arithmetic
Ten mailboxes at a conservative daily limit beat two mailboxes at five times the limit — same volume, a fraction of the risk.
SEC.05
List hygiene
>
Bounces are the fastest way to lose a domain. Providers read a high bounce rate as proof you are sending to a scraped list, and the penalty outlasts the campaign. Verify every address immediately before the send, not at the point you bought the data.
-Verify at send time — contact data decays continuously
-Drop catch-all and role addresses unless you have a specific reason to keep them
-Suppress unsubscribes, complaints and hard bounces globally, across every campaign
-Never re-import an old list without re-verifying it
SEC.06
Content and replies
>
Once identity and reputation are healthy, content decides the rest. Cold mail that reads like a newsletter gets filtered like one. Keep messages short, plain-text in feel, and specific enough that the recipient can tell it was written for them — genuine relevance is the only reliable engagement strategy.
✕Image-heavy templates and tracking pixels in the first touch
✕Multiple links, shorteners, or attachments before a reply
✕Identical copy across thousands of recipients
✕Follow-up sequences that continue after someone replies or opts out
SEC.07
What to monitor
>
Deliverability fails quietly: reply rate sags for two weeks before anyone checks placement. Watch the leading indicators per mailbox and per domain, and pause on the metric rather than on the anecdote.
watchlist.md
- bounce rate
- rising bounces = list or reputation problem
- reply rate
- the truest placement proxy you have
- complaint rate
- treat any sustained rise as an emergency
- per-mailbox trend
- one bad mailbox drags a whole domain
- DMARC reports
- catches unauthorised or misconfigured senders
>
Orbis runs these controls as infrastructure — authenticated warm mailboxes, enforced daily limits, verification before every send, and automatic pausing when a metric turns. See warm infrastructure or pricing.
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
>
Related: how to build an ideal customer profile.
[boot]EOF — © 2026 SOLVENT GLOBAL LTDok
$ready▍
● readylines 30/30
4398 B