Email courses: the complete guide to writing and finishing one

What an email course is

An email course is a fixed sequence of emails, written once, delivered one per day in a set order, to each reader individually, starting from the day that reader confirms their address. Nothing in it is scheduled to a calendar date: someone who joins in March and someone who joins in November get the same day one. The sequence has a first email, a last email and a defined number in between, and that number is published before anyone signs up.

Three properties separate it from everything else that arrives by email. It is finite: there is a last day, and the reader knows which one it is. It is ordered: day three assumes day two, and nobody gets them out of sequence. And it is relative to the reader rather than to you: the clock starts on their confirmation, which means the course keeps working at three in the morning on a public holiday while you are not thinking about it at all.

That last property is the one that makes the format economically different from writing. A newsletter is a standing obligation you pay weekly for the rest of its life. A course is a burst of writing and then a piece of infrastructure. The cost is front-loaded and finite, and the delivery is continuous and automatic, which is why a course written in one week can still be recruiting readers two years later without a further hour from you.

Why a finite sequence completes and an open-ended one does not

Subscribing to something open-ended is a decision made with almost no information. The reader does not know how often it arrives, how long each one is, how long it will keep arriving, or whether the fourth will be as good as the first. The only way to find out is to let a stranger into the inbox where their bank sends statements. A course removes every one of those unknowns at once. Five emails, one a day, here is what each covers, it ends on Friday. You are not persuading harder; you are asking for less.

The same boundedness works at the other end, where it is less obvious and matters more. Something with no ending can only be left by unsubscribing, which is a small negative act people postpone by quietly not opening anything. A course ends by being finished. Day five is a genuine event: the reader completes something, and a person who has completed something is on entirely different terms with you than a person who has stopped opening your mail without ever deciding to.

The daily rhythm compounds inside that window. Someone who opened day two is markedly more likely to open day three, because the sender name in the From line has stopped being unexpected and the subject line tells them exactly where they are in something they can see the end of. No other format earns that kind of familiarity in a single week, and the week is short enough that the reader never has to decide whether they still want it.

Choosing the promise, then the length

Start from a promise, not a subject. A subject is a territory: pricing, sourdough, container security. A territory can absorb an unlimited number of emails and therefore never tells you when to stop. A promise is one sentence with an outcome and an owner in it: by Friday you will have set your day rate and be able to say it out loud. The test is that a stranger can tell before signing up whether it is for them, and can tell on the last day whether you kept it. If either half is fuzzy, the course will be fuzzy, and the fuzziness shows up as an abandoned draft on day four.

Length follows from the promise rather than from preference. Count the distinct moves required to get a reader from where they start to where the promise puts them. If there are five, write five days. If the fifth move genuinely splits into two, you have six or seven. What does not work is choosing a length first and then dividing the material to fit it, because the days that exist to fill a slot are exactly the days readers stop on.

Five is the default for a reason that is structural rather than aesthetic. It is one working week, it is short enough that a stranger can picture the entire commitment in one glance at an opt-in page, and it is long enough for the rhythm above to establish itself. Seven suits a method with a real seventh step and adds a weekend, which for a course read at work is two days of lower engagement in the middle of the sequence. Ten is the outer edge worth considering: past that you are asking for a fortnight of someone’s attention, which is a newsletter’s ask without a newsletter’s ongoing reason to say yes.

The shape of one day

Assume a phone, a commute and ninety seconds. That is the median reading condition for anything that arrives in the morning, and it sets the length: roughly three hundred to six hundred words is comfortable. Below about two hundred a day rarely says enough to justify the interruption. Above about eight hundred, on a phone, the scroll bar itself becomes an argument for reading it later, and later usually means never.

Open with the thing itself. A line that says welcome back to day three of the course spends the reader’s first line telling them something the subject line already told them. Start with the idea, the observation or the mistake, let the structure be implicit, and end with exactly one action small enough to be done before the coffee is finished. A course works by accumulating five completed small things, and a reader who has done four of them is a different person on Friday from a reader who has read four of them.

Give the whole sequence one call to action, repeated, rather than a different ask each morning. Five different asks in five days reads as a sales sequence wearing a course as a costume, and readers recognise it by about day three. The last day is the one exception: it is allowed a single honest invitation.

The opt-in page, and the confirmation that follows it

An opt-in page for a course has an easier job than one for a newsletter, and the way to waste that advantage is to hide the specifics. State the number of days, state that it is one per morning, state what each day covers, and state what the reader will be able to do at the end. The page is the whole offer and the reader should be able to evaluate it without scrolling twice. Ask for an address and nothing else: every additional field is a further reason to leave, and a first name is rarely worth the ten per cent of sign-ups it can cost.

Then confirm. Confirmed opt-in, which is the same thing most people mean by double opt-in, sends one message to the address asking the owner to click a link, and delivers nothing else until they do. It is not a conversion tax to be toggled off when the numbers look thin. It is the only mechanism that distinguishes an address its owner typed from an address a bot generated or a stranger mistyped, and on a shared sending domain one unconfirmed list damages the delivery of every other sender on it. An unconfirmed address receives exactly one message, ever.

Expect to lose some of them there. Somewhere between a fifth and a third of sign-ups never confirm, which is the range commonly cited in published double opt-in comparisons, and the loss is mostly people who mistyped, people who were never really interested, and people whose confirmation landed in a promotions tab. The ones you keep are worth more than the ones you lose, which is the trade the mechanism exists to make.

What the confirmation message has to do, in the order the reader meets it:

• Say what they signed up for, by name, so a mistyped address produces a confused stranger rather than a complaint.
• Put the confirmation link high, as a real link with readable text, not as an image and not as a bare URL.
• Say when day one arrives once they click, so the silence afterwards is expected rather than a failure.
• Carry a working unsubscribe and a real postal address: a confirmation request is still a commercial message.
• Contain nothing else. No lesson, no offer, no attachment. One job, one link.

One per morning, in the reader’s timezone

The delivery rule that makes a course feel like a course is one email per day, at roughly the same local hour, in the reader’s own timezone. A sequence that arrives at nine in the morning for readers in Amsterdam and at three in the morning for readers in Los Angeles is a different product for the second group, and the difference shows up as a lower open rate on every day after the first. If you cannot determine a reader’s timezone, a fixed hour in a stated timezone is honest and workable, but say which one on the opt-in page.

A send hour somewhere between six and nine in the morning local time is the conventional choice, and the reasoning is that the email is near the top of the inbox when the reader first looks at it. That is a convention rather than a law, and for a course aimed at a specific working pattern the right hour may be very different. What matters more than the exact hour is that it does not move: a course that arrives at seven on Monday and at two on Tuesday has broken the one promise its format makes.

Day one is the exception, and it should arrive immediately on confirmation rather than waiting for the next morning window. The reader has just clicked a link and is looking at their inbox; that is the highest-attention moment the course will ever get and holding day one back for fourteen hours spends it. From day two onward, the daily window applies, counted from the day of confirmation rather than from a calendar.

What completion rates to expect, and why they are click-asserted

Plan for one in ten to one in four confirmed readers clicking the link on the last day of a five-day course, with longer sequences shading lower. Automated sequences click well above broadcast rates in the published benchmarks. The spread has almost nothing to do with the writing and almost everything to do with how well the promise matched the person who signed up: a course whose opt-in page was honest about who it was for lands near the top, one recruited broadly lands near the bottom, and the same emails are responsible for both.

Measure that with a click, not an open. Apple Mail Privacy Protection loads remote content, including the tracking pixel that an open is made of, through a proxy regardless of whether the reader ever looked at the message. On a list with a large share of Apple Mail readers, which is most consumer lists, the open rate is therefore a number that goes up when nobody reads anything. It is not merely noisy; it is systematically wrong in a direction that flatters you.

A click is an assertion: somebody chose to press something. So build each day around one link that a reader who actually did the thing has a reason to press, and define completion as a click on the last day. That number is several times smaller than an open rate and it is real, which makes it the only one worth reporting to yourself. It also survives the next privacy change, because no plausible mail client clicks links on the reader’s behalf.

The day after the course ends

A course that stops dead on the last day wastes the best asset it produced. On the morning after day five you have people who read five things you wrote, finished them, and are still opted in. Leave them with nothing and in three weeks your name in a From line is unfamiliar again.

There are only three honest answers to what happens next, and choosing one before you publish is part of writing the course. Hand finishers to a newsletter, which is the arrangement that works best for most people, because the course is the small legible ask and the newsletter is the open-ended one they are now in a position to judge. Or make one offer, once, on the last day, with the price and the shape of it stated plainly. Or say the course is over, and mean it: an explicit ending with nothing after it is a perfectly respectable product, and it is much better than a silence that becomes a sporadic broadcast six months later.

What does not work is deciding afterwards. A list assembled for a course and then mailed about something unrelated months later earns its complaints, because that is not what anyone agreed to.

How courses actually fail

Almost none of the common failures are failures of writing. They are failures of decision, and they happen before or after the drafting, which is why they are so hard to see from inside a document. The pattern is consistent: a promise too wide to keep, a length chosen before the material was counted, or a sequence that was never read on a phone in a real mail client.

The most expensive one is silent. Everything after the confirmation click is automatic, so anything broken there is broken identically for every reader and nobody is watching it happen. Send yourself through the whole course, from the opt-in form, on a real device, and read all five days as a reader gets them before anyone else does.

The failures worth checking for by name, roughly in order of how often they occur:

• A promise that is a territory rather than an outcome, which produces a course with no last day and a draft abandoned on day four.
• A recap of yesterday at the top of every email, which pushes the actual content below the fold for readers who did not need it.
• A different call to action each day, which is the tell that turns a course into a sales sequence in the reader’s mind.
• An email that depends on having clicked a link in an earlier day. Some readers did not, and a course that punishes them loses them entirely.
• Content rendered as an image, which is invisible to anyone reading with images off and unfindable in a mailbox search later.
• A course written against a price or a screenshot, which goes quietly wrong while nobody is reading it. Put a review date in the calendar the day you publish.