Email course and deliverability glossary

Alignment

Also called DMARC alignment, identifier alignment.

What is DMARC alignment?

Alignment is the DMARC requirement that the domain a reader sees in the From header match a domain that passed SPF or DKIM, which is why a message can pass both of those checks and still fail DMARC.

Mail passes SPF and passes DKIM and fails DMARC when both passes were on the sending provider’s domain while the From line said yours. That is the default state with a good many sending services until the domain setup is finished, and it looks identical to success everywhere except in a DMARC aggregate report.

SPF aligns on the envelope sender, the address bounces are returned to, and DKIM aligns on the domain named in the signature. Only one of the two has to align for DMARC to pass, and relaxed alignment accepts a subdomain of the same organisational domain, which is why sending from a subdomain reserved for mail breaks nothing.

Bulk sender requirements

Also called Google and Yahoo sender requirements, February 2024 email requirements.

What are the bulk sender requirements?

The bulk sender requirements are the list Google and Yahoo both began enforcing in February 2024: SPF and DKIM on the sending domain, a DMARC record on the From domain, one-click unsubscribe in the headers, unsubscribe requests honoured within two days, and a spam complaint rate under 0.3 per cent.

Google defines a bulk sender as one sending close to 5,000 messages a day or more to Gmail addresses from the same domain, and applies the stricter items above that line. The authentication floor applies to everyone at any volume: unauthenticated mail is refused rather than filtered, and Gmail returns a 5.7.26 error for it.

The practical reading for a small sender is that the whole list applies anyway. You will cross the bulk line on a good month, none of it is work worth doing reactively, and between them the two providers hold most of the consumer addresses an English-language course is sent to.

Call to action

Also called CTA.

What is a call to action in an email?

A call to action is the single thing an email asks the reader to do, stated plainly and linked once, which in a course email is usually the step that day’s lesson was teaching rather than a request to buy anything.

One per email is the working rule. A message with three asks is a message with none taken, and in a teaching sequence the ask is also the measurement, because a click on a link that a reader who actually did the thing has a reason to press is the only progress signal worth counting.

Put the link where it belongs in the teaching, on a worksheet, the next step or a reply prompt, rather than at the bottom as a read-more. A course can afford a plain ask: the reader confirmed their address for this, and the sequence is the reason the message is in front of them.

Click-asserted progress

Also called click-based completion.

Click-asserted progress is measuring how far a reader got through a course by acts that required a decision from a person, such as a click on a link in the last lesson, rather than by an open, which is an image load a machine can perform.

The reason to define progress this way is that the alternative has stopped working. Apple Mail Privacy Protection loads remote content through a proxy whether or not the reader looked at the message, Gmail pre-fetches and caches images, and clients with images off register nothing at all, so an open rate is wrong in a consistent and flattering direction.

Plan for one in ten to one in four of confirmed readers clicking on the last day, with longer courses lower. Treat the figure as an upper bound rather than a headcount, because security scanners and link previewers press links too.

Completion rate

What is a good completion rate for an email course?

A completion rate is the share of confirmed subscribers who reach the end of a course, and for a five-day course measured by a click on the last day it commonly runs between one in ten and one in four, with longer courses lower.

It is worth measuring because it says whether the thing you wrote is being used, and because it is the one funnel figure that cannot be improved by putting more strangers on a list. A course with a high sign-up rate and a low completion rate is a promise that was better than the material behind it.

Measure it with a click rather than an open, and give every day one link a reader who did the work has a reason to press. Published benchmarks disagree by wide margins here, because completion moves more with who the reader is and what was promised than with anything you can edit in the copy.

Confirmed opt-in

Also called double opt-in, DOI.

What is confirmed opt-in?

Confirmed opt-in, usually called double opt-in, is the practice of sending one message to a newly submitted address and adding it to a list only after somebody clicks the link in that message.

The word “double” is a misnomer and it has caused real confusion, because it sounds like a second form or a second act of typing. It is one extra click, in a message the reader was expecting, within seconds of asking for it.

It costs a share of sign-ups, typically a fifth to a third, and the loss is concentrated in the addresses you would not have wanted: the mistyped, the borrowed, the reluctant and the automated. What it buys is a list on which every address was reachable and somebody at it asked for the mail, which is what holds a complaint rate below the level the providers enforce.

A confirmation click is also the most durable consent record available. A form submission proves that a browser sent you a string; a click in a message delivered to the address proves that whoever reads that mailbox wanted this.

Consent record

Also called proof of consent.

What is a consent record?

A consent record is the evidence that a particular person agreed to receive a particular kind of mail, kept because Article 7(1) of the GDPR puts the burden of demonstrating consent on whoever relies on it.

What to keep is what you would need in order to answer a complaint a year later: what the person was shown, what they agreed to, when, and from where. A timestamp and an address on their own describe a row in a table rather than an act by a person.

A confirmation click is the cheapest version of this, because it is a timestamped act performed at the address itself. The GDPR names no mechanism and does not contain the phrase double opt-in; it defines consent as freely given, specific, informed and unambiguous in Article 4(11), and leaves the evidence to you.

Deliverability

Also called inbox placement.

What is email deliverability?

Deliverability is the share of the mail you send that reaches the inbox of the person you addressed it to, as distinct from delivery, which is only the share the receiving server accepted rather than bounced.

A message can be delivered and never seen: accepted at the door, filed straight into the spam folder, counted as a success by the sending side and read by nobody. The receiving provider does not report which folder it chose and will not, which is why deliverability is inferred from complaint rates, bounces and behaviour rather than read off a dashboard.

For a small sender it comes down to authenticated mail on a domain you control, a spam complaint rate under the level the large providers enforce, hard bounces suppressed on the first failure, and a list that confirmed it wanted the mail.

Delivery rate

Also called accepted rate.

What is a delivery rate?

A delivery rate is the share of a send that the receiving mail server accepted instead of bouncing, and it says nothing at all about whether any of it reached an inbox.

It is the number a sending dashboard can actually see, which is why it is the number sending dashboards show. Ninety-nine per cent delivered is entirely compatible with every message sitting in a spam folder, so a sender can look healthy on their own reports for months while their mail quietly stops arriving anywhere a person looks.

The useful thing to do with it is watch it move rather than admire it. A drop means bounces, and a jump in bounces means either a list nobody confirmed or a suppression list that is not being applied before the send.

Disposable email domain

Also called throwaway address, temporary email.

What is a disposable email address?

A disposable email domain is one that hands out addresses which expire or forward nowhere after a short period, typed by people who want the thing being offered without the mail that comes with it.

They are not a deliverability disaster in themselves, because a confirmation step removes most of them for free: an address nobody intends to read is an address nobody clicks a link at. What they cost is list size that was never real, and a completion rate calculated against people who were never there.

Blocking them by name is a maintenance job with no end, since the list of such domains changes weekly and every blocklist catches ordinary providers by accident. Confirmation is the cheaper control and it needs no list.

DKIM

Also called DomainKeys Identified Mail.

What is DKIM?

DKIM is a cryptographic signature over a message’s headers and body, made with a private key held by whoever sends the mail and checked by the receiver against a public key published in DNS under a selector, and it is defined in RFC 6376.

It survives forwarding, which is exactly where SPF fails, because the signature travels with the message rather than describing the route it took. The two are complementary rather than alternatives, and the bulk sender requirements ask for both.

The signature names a domain of its own, and that domain is what accumulates reputation. If your provider signs with its own domain rather than yours, the reputation being built belongs to the provider and to every other sender in its pool.

A DKIM key cannot be discovered. The public key sits at selector._domainkey under your domain and the selector is an arbitrary string chosen by whoever set it up, so there is no way to enumerate the keys on a domain; any signed message you have received names the one in use in the s= tag of its DKIM-Signature header.

DMARC

Also called Domain-based Message Authentication, Reporting and Conformance.

What is DMARC?

DMARC is the DNS policy that ties SPF and DKIM to the domain a reader actually sees, telling receivers what to do with mail from your domain that aligns with neither and asking for aggregate reports about it, and it is defined in RFC 7489.

The record lives at _dmarc under your domain, and its policy tag is the point of it: none means report what is happening and deliver it anyway, quarantine means treat failures as suspicious, and reject means refuse them. A domain that sends no mail at all should publish a reject policy, which is the cheapest anti-impersonation measure available to anybody who owns a domain.

A policy of none is the right place to start and the wrong place to stay. Its purpose is to collect aggregate reports for a few weeks so you can find the legitimate senders you had forgotten about and fix their authentication; a domain that has been on none for three years has a record that satisfies the bulk sender requirement and has never once caused a forged message to be refused.

Domain warming

Also called IP warming, warm-up.

How long does it take to warm up a sending domain?

Domain warming is the practice of raising the volume sent from a new domain gradually, over something like two to four weeks, so that a domain with no reputation does not arrive looking like one registered this morning for a campaign.

Neither Google nor Yahoo publishes a ramp; both say only to increase volume gradually and watch the signals, so anyone quoting a schedule to three decimal places is quoting a habit. The shape most senders use is a few hundred messages a day to the addresses most likely to engage, held for two or three days, then roughly doubling every two to three days while complaint and soft bounce rates stay flat.

If a rate moves you hold rather than push through. A course sender gets part of this for free, because a course fans out one message per subscriber per day rather than one blast to everybody at once.

Drip campaign

Also called automation sequence, autoresponder.

What is a drip campaign?

A drip campaign is a stored sequence of emails sent to a person on a timer after some trigger, run towards a goal the sender chose, which is the same machinery an email course runs on and a different promise to the reader.

The distinction is not technical. Both are a stored message, a rule about when it goes out, a queue and a suppression list; a course tells the reader what arrives on each of a fixed number of days, and a drip campaign is a marketing sequence with an objective the reader was never told about.

A course assembled inside a builder designed for drips inherits defaults nobody chose: a clock counted from the last send rather than from a morning, a trigger built to fire more than once, a goal that pulls people out early, and an unsubscribe that belongs to a workflow rather than to a person.

Email course

Also called email mini-course, course by email.

What is an email course?

An email course is a finite sequence of lessons delivered one email at a time, in order, on a clock that starts when each reader joins, and it ends on a day both of you knew about in advance.

Finite is the load-bearing word. A newsletter is a standing arrangement with no end and a course is a promise with a last day, which is why people finish courses and do not finish newsletters, and why a stranger will agree to five mornings when they would not agree to something indefinite.

The per-reader clock is the other half. Five messages scheduled a day apart work once, for the people already on the list; everybody who arrives on Wednesday starts at day three of something whose beginning they never saw.

Evergreen and cohort courses

Also called evergreen course, cohort course.

An evergreen course starts on the morning after each person confirms and is therefore running continuously, while a cohort course starts on one fixed date for everybody, so nobody in an evergreen course is ever late and nobody in a cohort is ever alone.

Both work, and the choice is between a shared moment and a thing that keeps working while you do not. A cohort earns its cost when the shared moment genuinely adds something, which in practice means a live element, a group, or a leaderboard that only makes sense if everybody is on the same day.

The usual mistake is running a cohort because a launch feels like an event. Absent a live element, a cohort is a deadline you have imposed on yourself, and every reader who finds the page the day after it starts is asked to wait or to leave.

Feedback loop

Also called FBL, complaint feedback loop.

What is an email feedback loop?

A feedback loop is an arrangement in which a mailbox provider reports back to a sender the messages its users marked as spam, so that those addresses can be suppressed immediately rather than mailed again.

Not every provider offers one and the terms differ: some report the complaint with the address and some strip it, and Gmail runs no per-message loop of the ordinary kind at all, offering aggregate rates in its postmaster dashboard instead. Most sending platforms enrol in the loops that exist and act on them without being asked.

What a loop is for is the one thing worse than a complaint, which is a second message to somebody who already complained. An address that arrives through a feedback loop belongs on the suppression list the same day, and it never comes back off it.

Five-day challenge

Also called 5-day challenge.

What is a five-day challenge?

A five-day challenge is a sequence of five daily emails that each ask the reader to do one small thing, usually with a sixth email that makes an offer, so the unit of progress is an action rather than a lesson.

It differs from a course in what finishes it. A course is finished when the reader understands something; a challenge is finished when the reader has done something five times, which makes the writing the smaller half of the job.

The design problem is what you ask for on each of the five mornings, and whether an ordinary person with a job can actually do it before lunch. An action that takes an hour is a challenge with a completion rate of almost nothing, however good the email asking for it was.

Hard bounce

What is a hard bounce?

A hard bounce is a permanent refusal of a message by the receiving server, carried in a 5xx SMTP response, because the mailbox does not exist, the domain does not exist, or the address has been disabled.

It must be suppressed immediately and permanently, on the first occurrence, not after three and not after a review. Continuing to send to an address you have been told does not exist is the clearest signal available to a receiving provider that a sender does not process its bounces, and not processing bounces is the operational signature of a bought or scraped list.

It is also how you keep mailing a spam trap, which is an address that exists only to be mailed by people who should not have it. A confirmation step removes most of the addresses that would hard bounce before they are ever sent to a second time.

Idempotency key

Also called Idempotency-Key header.

What is an idempotency key?

An idempotency key is a value a sender attaches to a request so that a provider receiving the same request twice performs the work once, which is what makes retrying after a timeout safe rather than reckless.

It matters in sending because the dangerous failure is not an error, it is silence. A request that times out may have been received, acted on and lost on the way back, so a naive retry delivers a second copy of a lesson to somebody who already has it.

The key has to be derived from the thing being sent rather than generated per attempt, and it has to be byte-identical on every attempt forever, or the second attempt is simply a different request. A row identifier is the usual choice, because it exists before the first attempt and never changes afterwards.

Lead magnet

Also called opt-in incentive, freebie.

What is a lead magnet?

A lead magnet is a specific, useful thing offered in exchange for an email address and permission to send more, and both halves matter: without the useful thing you are asking a stranger for their address for nothing, and without the permission you have a download page and no list.

It is not a sample of your paid work and it is not a brochure with a form in front of it. It is a complete small thing that solves one problem the reader actually has and leaves them better off even if they never buy anything, which is mechanism rather than generosity: a magnet only worth having if you buy the next thing is an advertisement, and it is opted into at the rate advertisements are.

The second thing to be clear about is what it is for. A lead magnet is not for downloads, it is for a list of people who open your mail, and those two goals pull in different directions, because the highest-converting offers tend to be the most generic ones.

List hygiene

Also called list cleaning, sunsetting.

What is list hygiene?

List hygiene is the work of keeping a list to addresses that exist and belong to people who still want the mail: hard bounces suppressed on the first failure, long-silent addresses retired, typos caught at the form, and nothing re-imported that was already suppressed.

The sunset window is the part that needs a number, and the honest answer is a range: most senders stop after somewhere between three and six months with no measurable engagement. A course sender can be stricter than a newsletter, because a course either gets finished or it does not.

What matters more than the exact window is that there is one and that it applies automatically. Hygiene done by hand is hygiene done when somebody remembers, which is after the complaint rate has already moved.

List-Unsubscribe

What is the List-Unsubscribe header?

List-Unsubscribe is a message header offering machine-readable ways out of a list, and paired with a List-Unsubscribe-Post header carrying the value List-Unsubscribe=One-Click it is what a mail client uses to unsubscribe a reader without opening a browser.

Two headers do the work and one-click requires both. List-Unsubscribe on its own predates RFC 8058 and offers exits a client may choose to present; List-Unsubscribe-Post is the header that declares the https URI safe to POST to without asking the reader anything further. A message with the first and not the second does not have one-click unsubscribe, whatever the sending dashboard says.

The header value usually carries two exits in angle brackets, an https URI and a mailto. The https one is what one-click uses, and the mailto is the fallback for clients that prefer it.

Mail Privacy Protection

Also called MPP, Apple Mail Privacy Protection.

What is Apple Mail Privacy Protection?

Mail Privacy Protection is the Apple Mail feature that loads a message’s remote content through a proxy whether or not the reader opened it, which is what stopped an open rate from measuring anything about a person.

Apple documents the effect plainly: the feature stops senders knowing when mail has been opened. On any list with a substantial share of Apple Mail readers, and most consumer lists have one, the open rate rises when nobody reads anything, so it is not noisy data but wrong in a consistent and flattering direction.

It also hides the reader’s address from the fetch, so the location and time-of-open inferences drawn from a tracking pixel are describing a proxy. What is left are the signals that require a decision by a person: a click, a reply, a confirmation.

One-click unsubscribe

Also called RFC 8058 unsubscribe.

What is one-click unsubscribe?

One-click unsubscribe is a mail client operating the unsubscribe on the reader’s behalf by POSTing to an https URI the sender put in the message headers, with no browser, no confirmation page and no further interaction from the reader, as RFC 8058 defines it.

There is no “are you sure”, no preferences screen, no cookie, no session and no login, and whatever the endpoint returns is thrown away. The correct response is a bare 200 with an empty or trivial body, and no redirect at all.

The redirect is the failure worth stating twice, because the reflexive thing to do in any web framework is to send the reader to a friendly confirmation page, and a redirect here fails the specification while looking completely healthy in every test a developer runs by hand. A browser follows it and shows a nice page; the mail client that posted the request is not reading the response.

The endpoint accepts the POST unauthenticated, so the token in the URL is the only thing identifying the subscriber. Make it long, random and single-purpose, and make the endpoint idempotent, because scanners will post to it more than once.

Open rate

Why is open rate unreliable?

An open rate is the share of delivered messages in which a tiny remote image loaded, used as a proxy for a human having looked, and it is unreliable because that image is fetched by proxies and caches that are not readers and never fetched by clients that block images.

The false positives and the false negatives arrive in proportions that differ by provider and change without notice. Gmail fetches and caches remote images through its own servers, Apple Mail Privacy Protection fetches them on the reader’s behalf whether or not the message was opened, and any client with images off registers nothing at all.

A sender who prunes a list by open rate is therefore pruning partly at random. There is a second objection with nothing to do with accuracy: the pixel puts a third-party host and a remote image load into every message, and an invisible image reporting a reader’s behaviour is processing personal data that has to be disclosed and justified like any other.

Opt-in page

Also called landing page, sign-up page, squeeze page.

What is an opt-in page?

An opt-in page is a page whose only job is to describe one thing and collect an email address for it, and a dedicated one commonly converts 15 to 30 per cent of its visitors where an inline form in a sidebar converts 1 to 3 per cent.

The difference is not persuasion, it is that the page has one subject and one field. A visitor who arrived for the thing being offered is asked for one decision, and everything else on the page is either evidence for that decision or in the way of it.

Say what arrives, how many days it runs and what happens next, because the sentence on this page sets the expectation the confirmation email then has to meet. A reader who does not recognise what they signed up for eleven seconds later is where a complaint comes from.

Preheader

Also called preview text, preheader text.

What is a preheader in an email?

A preheader is the line of text an inbox shows after the subject when there is room, taken from the top of the message body, and it is a separate opportunity rather than an extension of the subject line.

On a phone it is often the line that decides, and left unset it fills with whatever the email happens to begin with, which is frequently a logo alt text or the words “View this email in your browser”.

Writing it is close to free and almost nobody does it. In a course it is the natural home for the day number, on the mornings when the subject line needs its whole width for something more interesting.

Re-permission campaign

Also called reconfirmation campaign, permission pass.

What is a re-permission campaign?

A re-permission campaign is a single message sent to an existing list asking the people on it to confirm that they still want the mail, after which everybody who did not confirm is suppressed rather than mailed again.

It is what you run when you cannot show how a list was collected, when it has been silent for a long time, or when it arrives with a business you acquired. It is expensive in list size, and it is the only way to turn a list you cannot account for into one you can.

The message has to stand on its own, because most of the people receiving it do not remember you. Name the thing they signed up for, say when and where if you know, say what happens if they do nothing, and ask for exactly one click.

Send it in batches rather than all at once. A large send to a list with a poor complaint rate is the send that damages the domain you were trying to rescue.

Role address

Also called role account, generic address.

What is a role address?

A role address is one that belongs to a function rather than to a person, of the info, support or sales kind, and it is read by several people, by nobody, or by a ticket system.

Mailing one is riskier than mailing a person for two reasons. Nobody at a shared mailbox has personally agreed to anything, so a complaint is likelier; and several of the addresses used as spam traps are role addresses, precisely because a role address that never signed up for anything is a good way to catch a sender who did not ask.

A confirmation step settles this without a blocklist. Somebody has to open the mailbox and click, which is exactly the evidence a role address otherwise cannot give you.

Seed list

Also called seed testing, inbox placement test.

What is a seed list?

A seed list is a set of addresses you control at several mailbox providers, mailed alongside a real send so that you can open each account yourself and see which folder the message landed in.

It exists because no provider reports inbox placement back to a sender, so the only way to observe the difference between delivery and deliverability is to be the recipient. A handful of accounts at the providers your readers actually use tells you more than any score a tool will sell you.

Treat the result as an indication rather than a measurement. Seed accounts have no reading history, no contacts and no engagement, and filtering is personalised, so a message that reaches a seed inbox can still be filtered for a reader who has never opened one of yours.

Sender reputation

Also called domain reputation, IP reputation.

What is sender reputation?

Sender reputation is what a mailbox provider has learned about the domain and the infrastructure your mail comes from, assembled from complaints, bounces, spam trap hits and how the people who receive your mail behave, and it is what decides the folder.

It is held per provider and it is not portable: Gmail knows nothing about what Yahoo thinks of you. Nobody publishes the function, and the only window most senders get is the provider’s own postmaster dashboard, which is free to whoever proves they control the domain.

Now that authentication is a floor, the identity carrying reputation is the domain in your DKIM signature rather than an IP address. That is the argument for sending from a subdomain reserved for mail: a burned subdomain can be retired in an afternoon, and a burned root domain takes your invoices and your password resets with it.

Sending domain

A sending domain is the domain that signs and sends your mail, and it is worth separating from the sending infrastructure: the IP addresses mail leaves from are shared by default at every provider serving small senders, while the domain in the signature and the From line is yours alone.

The decision that matters for a small sender is therefore not shared against dedicated IP addresses. IP reputation is built by sustained volume, and a dedicated IP under a low-volume sender is an IP with no reputation at all, which is worse than a share of a well-run pool.

What there is to decide is which of your domains does the sending. Use a subdomain reserved for it and keep the root domain for the mail humans send each other, because a burned subdomain can be replaced in an afternoon and a burned root domain cannot be replaced at all.

Single opt-in

Also called SOI.

What is single opt-in?

Single opt-in means an address starts receiving mail as soon as it is typed into a form, with nothing in between to establish that the person who typed it owns the mailbox it points at.

Its advantage is real and it is the reason it persists: nobody is lost between the form and the list. Its cost is that an email form accepts any string shaped like an address, so a share of what arrives is a typo, a colleague, somebody who wanted the download and not the mail, or a bot.

From a mailbox provider’s point of view every one of those is indistinguishable from a bought list. They bounce, or they sit and never engage, or they press the spam button, and all three are measured against you.

Soft bounce

What is a soft bounce?

A soft bounce is a temporary refusal of a message by the receiving server, carried in a 4xx SMTP response, because the mailbox is full, the server is greylisting you, or you are being rate-limited.

It is retried with backoff rather than suppressed, and reclassified if it keeps happening. There is no published figure for when a soft bounce becomes permanent, because it depends on the reason code, but a common shape is to retry over a day or two and then treat somewhere between three and seven consecutive soft failures across a week or two as a hard bounce.

A mailbox that has been full for a fortnight is not a reader. A rise in soft bounces at one provider and nowhere else is also worth reading as a message about you rather than about the mailboxes, because that is what throttling looks like from the sending side.

Soft opt-in

Also called existing customer exemption.

What is the soft opt-in?

The soft opt-in is the narrow exception in European electronic marketing rules that lets a seller market its own similar products to somebody whose details were obtained in the course of a sale, provided that person was given a clear chance to object when the details were collected and in every message since.

It lives in the ePrivacy Directive as implemented in national law, and in the United Kingdom in the Privacy and Electronic Communications Regulations, rather than in the GDPR. That is why it differs between member states, and why it is the one place where being in one country rather than another genuinely changes the answer.

Each condition is read narrowly: your own products, similar ones, a real sale or negotiations for one, and an opt-out offered at collection. A list of people who bought something from you is not automatically a list you may send a course to.

Spam complaint rate

Also called complaint rate, abuse rate.

What is a good spam complaint rate?

A spam complaint rate is the count of people who pressed the spam or junk button divided by the mail delivered to that provider over a window it chooses, and Google and Yahoo both name 0.3 per cent as the level to stay under and 0.1 per cent as the level to aim at.

The gap between the two figures is not fussiness. The higher one is where filtering starts to be applied to you, and the lower one is the headroom that means a single bad send does not take you across it.

Because these are rates they behave worse the smaller you are: at a thousand delivered messages three complaints is already 0.3 per cent, and at two hundred a single complaint is 0.5 per cent. A small sender has no statistical cushion, only a structural one.

Complaints are not really about content. Somebody presses the button when they do not recognise the sender, cannot find a way out, or are surprised by what arrived, which points at an address added without its owner asking, a long gap before the first send, a frequency nobody promised, and an unsubscribe that is hidden, slow or broken.

SPF

Also called Sender Policy Framework.

What is SPF?

SPF is a DNS record listing the servers allowed to send mail for a domain, checked by the receiver against the envelope sender rather than against the From line a reader sees, and it is defined in RFC 7208.

Two consequences follow and both get missed. It proves something about the route rather than about the message, and it breaks on forwarding, because the forwarding server is not in your record, which is exactly the case DKIM survives.

A domain may publish only one record beginning v=spf1. Two is a permanent error by the specification and the correct behaviour for a receiver is to fail the check outright rather than pick one, which is the most common way a domain breaks SPF for itself: a second provider’s record is pasted in alongside the first instead of its include mechanisms being merged into one.

Every include, a, mx, ptr, exists and redirect mechanism costs a DNS lookup, and the specification caps the total at ten, counted recursively through everything your includes pull in. Past ten a receiver returns a permanent error, so a record that worked last month can stop working because a provider added an include inside their own.

Subject line

How long should an email subject line be?

A subject line is the line an inbox shows beside your name, and the length that matters is not the total but the front of it: put what is interesting in the first 30 to 40 characters, which is roughly what an iPhone held upright will show.

An inbox truncates at a pixel width in a proportional typeface rather than at a character count, so every figure quoted in characters is a translation that moves with the letters you chose. Outlook on a desktop shows roughly 60 and Gmail in a wide browser window roughly 70, and the shortest of the three is your working budget, because it is the one your reader is most likely holding.

A course subject line does not have to sell anything, which is a freedom worth using. The reader confirmed their address for this and the sequence is the reason the message is there, so a plain description of the day beats anything clever, and a “Day 2 of 5:” prefix costs about eleven characters for a real gain in orientation.

Suppression list

Also called do-not-mail list, exclusion list.

What is a suppression list?

A suppression list is the permanent record of addresses that must never be mailed again, holding unsubscribes, complaints and hard bounces, and it is checked before every send rather than after one.

It is the most valuable list you own and the one most often lost. It has to survive every re-import, every change of sending platform and every export to a spreadsheet, because the way somebody gets mailed after they left is almost always a list that went around a suppression check rather than through it.

When somebody exercises a right to erasure you still need to not mail them again, which is the apparent contradiction at the centre of this. It is usually resolved by keeping the minimum necessary to honour the request rather than the record itself: a one-way hash of the normalised address, with no name, no history and no profile attached, so what is retained is a prohibition rather than a subscriber.

Throttling

Also called rate limiting, deferral.

What is throttling in email sending?

Throttling is a receiving provider deliberately slowing or temporarily refusing mail from a sender, usually with a 4xx response, because the volume, the pattern or the reputation it is seeing is more than it is willing to accept right now.

From the sending side it looks like a rise in soft bounces and deferrals at one provider and nowhere else, which is the tell. The correct response is to send more slowly and find out what changed, not to retry harder: a sender that answers a deferral with more connections is confirming the reason it was deferred.

A course fans out one message per subscriber per day rather than one blast to everybody, which spreads volume naturally and is one of the quieter advantages of the format.

Unsubscribe

Also called opt-out.

How quickly must an unsubscribe be honoured?

An unsubscribe is a reader withdrawing permission to be mailed, and since February 2024 Google and Yahoo have required one to be honoured within two days, a much shorter clock than the 10 business days the United States CAN-SPAM Act allows.

Two days is the outer limit and there is no reason to use it. An unsubscribe is a single row and the send is asynchronous, so the honest implementation removes the address before it answers the request; the senders who need the two days are the ones running an export to a spreadsheet and an import back, which is also how somebody gets mailed after they left.

Every commercial email carries two exits: the headers, for clients that can operate them, and a visible link in the body for everybody else. The link is not optional, because not every client renders an unsubscribe control, a forwarded message loses it entirely, and CAN-SPAM requires a clear and conspicuous explanation in the message itself.

A message that arrives after somebody asked to leave does not produce a second unsubscribe. It produces a spam button press from a reader who now believes you ignored them, and those are the complaints that move a complaint rate fastest.

Welcome sequence

Also called welcome series, onboarding emails.

What is a welcome sequence?

A welcome sequence is the short run of emails a new subscriber receives after joining a list, sent to introduce who you are and what will arrive, and it is the mail people are most likely to read because it is the only mail they are expecting.

It differs from a course in what it is for. A course teaches something and ends on a day the reader was told about; a welcome sequence sets an expectation for an ongoing arrangement, so it can be two messages or five and it has no natural last day of its own.

The first message is the one that matters, because it arrives while the reader still remembers signing up. Name the thing they asked for, say what arrives and when, and key the sequence to the moment they joined rather than to a date on your calendar.