Outbound email, sent from your domain and written by a person.
We write, design, send and follow up B2B outbound email, from your domain and your ESP rather than ours. Authentication is set up and verified before the first message goes out, volume stays inside what your domain and your complaint rate can carry, and you read every message before it is sent.
The cadence, as files rather than as a diagram.
Plain, typed, and sent from your domain. No images to load, no button to render, nothing that looks like a campaign. These three are in the free catalogue on this site and the HTML is yours.
- Right for you if
- You have a list you believe in and nobody with the time to write to it properly, or your sending has quietly stopped landing and nobody can say when it started.
- What it includes
- Copy, design, the sending domain and its authentication, ESP setup, the follow-up cadence, reply handling, and a weekly read of what actually replied.
- What it does not include
- We do not send from a throwaway domain bought for the purpose, we do not buy or scrape lists, and we will not push volume past what your domain and your complaint rate can carry.
The mailbox providers changed verb, so this is now the first week of work.
Non-compliant bulk mail used to be filtered, which meant it arrived somewhere you did not want it. Now it is rejected, which means it does not arrive. Enforcement has been fully active across Google, Yahoo and Microsoft since May 2026, with Google and Microsoft issuing permanent 550 rejections, and roughly 30% of bulk senders are still partially non-compliant on at least one requirement.
5,000
messages a day
The point at which a domain has to pass all of it rather than most of it. Counted across every system that sends as you, not just the one you are thinking of.
0.10%
spam complaints, the working ceiling
One complaint per thousand delivered. Google asks senders to stay below this, which on a list of ten thousand is ten people.
0.30%
spam complaints, never reach it
The number Google asks senders never to hit. Above it, the reputation damage outlasts the campaign that caused it.
The detailClose
Compliant senders average about 89% inbox placement, while non-compliant senders see 22% to 34% of their mail routed to spam. That gap is not a copywriting problem and no amount of subject line testing closes it, which is why we do this part before we write a word.
- SPF and DKIM that align, not merely pass
- Google's own guidance is that the domain in the From header must align with either the SPF domain or the DKIM domain. SPF passing for your ESP's bounce domain while failing to align with the domain your recipient reads is the most common broken state we find, and it looks healthy from every dashboard.
- DMARC published, then moved
- A record at p=none with a real reporting address on day one, because that is what turns on the aggregate reports naming every system that sends as you. The move to quarantine and then reject waits for two weeks of clean reports, because it is the only step that can break something.
- One-click unsubscribe that unsubscribes
- The RFC 8058 header pair, an endpoint that accepts an anonymous POST, no login screen behind it, and requests processed within two days. It is the cheapest complaint-rate insurance there is, because the alternative to an easy unsubscribe is the spam button.
- Complaint rate under the ceiling
- Google asks senders to keep spam rates in Postmaster Tools below 0.10% and to avoid ever reaching 0.30%. At 0.10%, one complaint per thousand messages, a program sending six hundred messages a month has no headroom inside the ratio at all. Precision is the deliverability strategy here, not an editorial preference.
- PTR, forward-confirmed DNS and TLS
- Usually your platform's job, verified anyway. These are refusals at connection time rather than spam-folder placements you can measure afterwards, so a failure here is invisible in every report you have.
What we do, and what stays with you.
The split matters more than the scope, because the split is the part people find out three weeks in. Here it is in advance.
We do
- We write
- The copy, the sequence and the follow-up cadence, built from your own words: sales calls, support threads, the emails that already worked. A first draft can come from a model. The version that leaves is edited by the person who will be on the reply.
- We design and build
- The email itself, built to render in the clients your list actually uses, and the landing page behind it where there is one.
- We set up the sending
- SPF, DKIM, DMARC, the branded sending domain at your ESP, suppression and unsubscribe handling, seed addresses, and a sending schedule that does not put everything on the same hour of the same day.
You keep
- You approve every message
- Nothing is sent that you have not read. This is not a courtesy. It is the control that keeps a wrong fact out of an email to a named account, and it is the reason the first send is week four or five rather than week one.
- Your IT publishes the records
- We write the exact values and a line saying what each one does, and your team puts them in. We do not ask for control of your DNS zone, and an agency that does is asking for something it does not need.
- You own the domain, the data and the replies
- Including the replies that say no, which are the most useful thing a program produces in its first quarter.
How replies are handled.
A reply is the only thing outbound produces that is worth anything, so the handling is agreed before the first send rather than improvised on the day one arrives.
A reply arrives
In your inbox, on your domain, to a person who exists.
Triaged that day
Sorted into interested, not now, wrong person, and stop.
The research note travels with it
Whoever picks it up gets the dated reason and the checked facts, not just the thread.
Drafted by us, sent by your named person
We write the answer to the objection. It goes out under the name that has been on the thread all along.
Yes
A call in the diary, with the note attached so nobody opens with a question the account has already answered.
No
Recorded with the reason, in words, against the account. That file is the most useful thing a program produces in its first quarter, and it is the reason the list changes.
The detailClose
Replies land in a mailbox on your domain that a named person on your side watches, and we watch it with them. Triage is daily. The interested ones go straight to whoever is taking the meeting, with the research note attached, so they walk in already knowing why the account was on the list. Objections are drafted by us and sent by the person whose name is on the email, because a reply from a different address ends the conversation. The nos are recorded with the reason, and the reason is what changes next month's list.
Out of office replies, wrong-person replies and role changes are all data. A wrong-person reply naming the right person is the highest quality signal the whole program produces, and it is the one most often thrown away.
We do not run an automated reply handler, and we do not let a sequencer keep sending to somebody who has answered. The follow-up stops the moment a human does.
The rest of it.
The parts that are an argument rather than a mechanism. Open whichever one you are actually weighing.
We send from your domain, not from a lookalike.
A lot of outbound is run from a secondary domain bought for the purpose, one character away from the real one. It is done to protect the sender's reputation, and the reputation it protects is the agency's. We will not do it.
Two reasons. The rules are evaluated against your organisational domain, so the address your invoices and renewal notices go out from is the one carrying the program's behaviour, and that is exactly the thing worth being careful with. And a named account that looks up the domain you wrote from and finds a three-month-old registration has already answered the question of whether you are worth replying to.
So the setup is your domain, your ESP, your DNS records and a volume ceiling set by what that domain can carry. If the honest ceiling is lower than the plan needs, we say so before the contract rather than after the first bounce report.
The operations underneath it: ESP setup and reporting.
The plumbing is part of this service rather than a separate line on the invoice, because a program nobody can measure honestly is a program that gets renewed on feeling.
ESP setup means the branded sending domain, the suppression list, the unsubscribe endpoint, the seed addresses, the templates in your account and the schedule. Reporting means a weekly note saying what was sent, what replied, what bounced and by which class, what unsubscribed and what the complaint rate did.
What that report does not lead with is an open rate. Apple's Mail Privacy Protection downloads remote content in the background by default, regardless of whether the recipient engages with the email, so a share of every open figure is a machine rather than a person. Replies, meetings, bounces and complaint rate are the numbers we will argue with you about.
What we do not do.
We do not run AI SDRs on a client domain. After AI adoption, per-rep monthly volume went from roughly 1,150 emails to 7,400 while raw reply rates fell from 4.7% to 2.9%, with about 1 in 6 emails never reaching the inbox. On raw reply count the volume play does win, and that is the part most critiques skip. It is simply not available here: a program of a hundred and fifty accounts contains perhaps seven hundred contactable people, so the only way to reach that volume is to send the same people the same thing more often, which is the single variable reply rate is most sensitive to.
We do not buy lists and we do not scrape. We do not take over your CRM. And we do not work to a volume target, because a volume target is a thing you pay for in domain reputation.
The long version.
We write these up properly, with the sources in line, so you can check the reasoning rather than take our word for it.
Questions.
Do you send from our domain or from yours?
Yours, with your ESP and your DNS records. Sending from a lookalike domain we owned would protect our reputation at the cost of yours, and the rules are evaluated against your organisational domain anyway, so the protection is largely imaginary.
Our domain has never sent cold email. Where do we start?
Authentication first, then deliberately small sends watched against the reports. Early volume goes to the most likely repliers, and the ceiling rises only while bounce rate and complaint rate stay flat. There is no published warm-up schedule we would stand behind, so we treat it as a measurement problem rather than a recipe.
How many emails a day will you send?
Fewer than you expect. The list sets the volume, not a target, and most account-based programs sit well under the five thousand a day threshold that pulls a sender into the bulk rules. We run the checks anyway, because rejection is evaluated across your whole domain and another department's newsletter shares your reputation.
Who answers the replies?
A named person on your side owns the mailbox and takes the meetings. We triage daily, draft the answers to objections, and route the interested ones with the research note attached. Nothing is auto-replied, and the sequence stops for anyone who has written back.
Will you report open rates?
As a health check, with the caveat next to it, never as a result. Apple downloads remote content in the background whether or not the recipient engages, so a meaningful part of any open number is a proxy rather than a reader. Replies, bounces and complaint rate survived intact and those are what we report on.
Can you fix our sending setup without running the campaigns?
Yes. That is the deliverability service, and it stands on its own. Plenty of teams want the records right and the complaint rate understood while continuing to write the emails themselves, and that is a sensible thing to want.
Sources on this page: Google, email sender guidelines, PowerDMARC, bulk email sender requirements, Red Sift, bulk email sender requirements, MarTech, new rules for bulk email senders, RFC 8058, one-click unsubscribe, Apple, Mail Privacy Protection, Salesmotion, AI SDR tools compared, 2026.
Start a conversation.
Tell us what you are trying to fix. If we are the wrong people for it, we will say so in the reply and point you at what we would do instead.
- A reply from a person, within one working day.
- If it sounds like a fit, a thirty minute call. No deck.
- A written plan within a week of that call, and no retainer until you have read it.
If a form is not your thing, hello@draftship.io reaches the same inbox.