Before Any Advice on Writing an Email Newsletter, the Sender and the Subject Line Have Already Settled Whether It Is Opened

The message written today is not what produced most of today’s figures. What settled the base of how it was opened is what has been sent over the past several months. The subject line moves the day’s portion on top of that. And today’s message is an input into how things get opened from next month onward.

That time lag is what makes improving an email newsletter difficult. Change the subject line and the day’s figures move. Change the substance and the day’s figures do not. Where the hand was put in and where the figure moved do not land in the same round.

Not landing together, no judgement is possible. When the figure rises in a month where the subject line changed, the reason may be the subject line, or it may be that a change of substance three months ago has begun to take effect. The two appear as the same figure.

And where judgement stays impossible, the criterion of selection swaps out from effect. Whatever produces a result the same day gets selected. Subject lines are written every time, so they can be improved every time, and every improvement leaves a record. The design of substance takes months to leave one. The swap occurs not as a decision but as a way of selecting work.

What circulates as advice is mostly about the text. How to set the subject line, how to open, how to build the structure, how to close. All of it works. It simply works only after the message has been opened.

When somebody decides whether to open, what is being consulted is the accumulated memory of everything that has arrived from that same sender. The information visible before opening is limited, and the one processed first is not the subject line.

What can be solved and what cannot divide sharply. What can be solved: where opening is settled; the mechanism by which history builds an order; and what actually works on the side of the text. The text below proceeds in that order.

Two things cannot be solved. What percentage counts as good does not get settled here. That figure shifts with how the list was built and how the measurement works, and there is nobody to compare against. Nor does how often to send. Frequency has both a floor and a ceiling, and those two are settled by the nature of what is being handled.

Where the loss lands can be set out first too. Improvement confined to the text has no end. How to set a subject line can be improved indefinitely. While the improving continues, the variables that do not move stay outside consideration. What is not inside does not rise as a candidate reason for things going badly. And since only the record of continued improvement remains, from outside it looks as though everything has been tried.

The Basics of Writing an Email Newsletter Are All Operations That Conclude Within One Message

The advice set out below works. Its range, though, closes inside a single message.

At the centre sits the subject line. The account that opening is settled by the subject line comes first, followed by touching the reader’s difficulty, making clear what is gained by reading, and fitting within the displayed character count. Displays vary by environment, though a guide of around fifteen characters is often given.

Next, the sender’s name. Make the display such that who it arrived from is clear. A company name, a brand name, or a person’s name — in a form the recipient can recognise.

Next, the opening sentence. Enter from a seasonal greeting or a topical item and build familiarity. Do not go straight into the substance; build a posture for reading.

Next, the structure. Settle what goes into one message and arrange it in a readable order. Do not run too long, and insert breaks.

Next, the target and the purpose. Settle in advance who it is for and what it is for. Without that, the contents scatter.

Last, the time and frequency of sending. Deliver in a band when the other person is reading. Frequency neither too high nor too low.

These are accurate. Subject lines do sway opening, and where the sender is unrecognisable it does not get opened. Where the opening is hard to read, what follows goes unread.

Line the advice up and the range of its object sits in a single place. Every piece is written as an operation concluding within one message. Today’s subject line, today’s opening, today’s structure. The object of improvement is confined to the message in front of you.

The range is not an error. Making each message better is genuinely necessary work.

It breaks where variables exist that improvement confined to one message cannot move. Opening is the representative case. What settles whether something is opened is only the information visible before opening, and at its centre are two things: the sender and the subject line.

The set shows no order either. In practice, what gets touched first is whatever costs least. Subject lines are written every time and so improved every time, while the setting of target and purpose is settled once and never touched. What is untouched stays as it was when settled. Whatever costs least getting touched first shows up whenever the set itself indicates no starting point. The side of where to begin is held by the design of a marketing funnel.

The target and purpose are supposed to be settled first, so the order is given. It is written.

There is no occasion for re-settling written anywhere, though. The state of readers changes with time, and so does the range you handle. A form settled once at the outset cannot handle change. Re-settling usually happens after the figures fall. Re-settle after the fall and the effect of re-settling takes still longer to appear.

What Is Visible Before Opening Centres on the Sender and the Subject Line

What settles whether the subject line gets read is the sender’s name.

What lines up first on the recipient’s screen is the sender’s name. The subject line sits beside or beneath it. In some environments, part of the opening text lines up as well.

The order of the advice runs the other way. The subject line comes first and the sender is touched on briefly afterwards. In the actual order of processing, the sender comes first.

The role the sender’s name performs is not identification. It is the work of pulling out where that name sits in that person’s order of importance.

Some people do not read an inbox from the top down. They look for names they know and skip the rest. On the skipped side, however good the subject line is, it does not function.

Attention has a confirmed property. When somebody selects one among several pieces of information arriving at once, the contents of the unselected side do not remain in awareness.

In 1959, the experimental psychologist Neville Moray used the then-current experimental form of playing different speech into each ear and having one repeated back, and showed one exception (Moray, 1959, Quarterly Journal of Experimental Psychology, 11(1), 56–60). Almost nothing about the unrepeated side can be reported, yet a person’s own name was noticed by some participants even when it flowed in from the side attention was not on.

What happens in an inbox is a selection of that form. Contents are not being judged and discarded; the contents of what is not selected never rise into awareness at all. And what tends to pass through as an exception is a name already inside the order of importance.

Work on the subject line and it might catch the eye without being in the order. It sometimes does.

The methods for catching it are limited, though, and they wear out. Insert a number, use brackets, imply urgency. Each works once and stops working on repetition.

The reason they stop is that the form creates information: “a sender who sends things that way.” Once that information exists, the form gets identified first.

And that identification sometimes works faster than the name. The shape of the subject line enters the visual field before the name is read. A sender remembered by form gets processed by form rather than by name. Change the form and the processing restarts, and the restarted processing also hardens into a form.

The Order of Priority Is Made by History

Where the sender’s name sits in the order is settled by past memory.

The contents of that memory are simple. What happened when something from this name was opened before.

It divides three ways. Useful, merely time spent, or not remembered at all.

The third is the hard one to handle. Opened, but nothing retained. Where that continues, the name lowers in the order as “a name with nothing inside it.”

From here one consequence emerges. The work of raising the open rate is performed not by today’s subject line but by the substance of each past message.

And that work carries a lag. Change the substance today and the order moves several messages later. For those several messages, the figures do not change.

Because of the lag, the effect of changing substance cannot be separated from the effect of changing the subject line. Subject lines take effect the same day; substance takes months.

The subject line may get changed again during that period. Once changes overlap, what worked can no longer be known.

None of this says the subject line must not be changed. Where content and subject line do not match, a betrayal occurs after opening.

Improving the subject line and improving the substance differ in the time over which they take effect, so moving both at once makes judgement impossible.

The difference in time works on priority too. Subject lines end as a day’s work and take effect the same day, while the design of substance has no end and takes effect late. Which gets done first tends to be the former.

And from this structure emerges the solution of lowering the frequency and sending only the good rounds. It is one solution.

Leave too long a gap, though, and the name itself stops being recalled. A name inside the order holds its position each time it is consulted. Where the period without consultation runs long, what disappears is not the position but the existence.

So frequency has both a floor and a ceiling. The floor is the interval at which a name is not forgotten; the ceiling is the interval at which substance can be held. Where those two do not intersect, the form of sending itself is what comes up for review.

Whether they intersect cannot be measured before the frequency is set. The interval at which substance can be held is unknown until several messages have actually been written. The first frequency chosen is therefore a provisional value. Forget that it is provisional and the frequency becomes fixed while the substance thins to match it.

The order is to measure first the interval at which substance can be held, and if it is shorter than the interval at which a name is forgotten, decide within that range. If longer, either change the form of sending or lower the volume per message. Both come before forcing the frequency up.

Lowering the Order Is Fast, Raising It Is Slow

When history builds an order, the speeds up and down are not the same. Falling is faster and rising is slower.

This asymmetry has broad support on the psychology side. Roy Baumeister and colleagues gathered observations across impression formation, learning, relationships and emotion, and reported the regularity that for events of the same magnitude, the bad works more strongly, more quickly and for longer than the good (Baumeister, Bratslavsky, Finkenauer & Vohs, 2001, Review of General Psychology, 5(4), 323–370).

Carried into delivery, it runs like this. One message opened with nothing inside it is recorded immediately. A good experience raising a name’s position, by contrast, needs several messages in succession. One good instance cannot be told apart from chance.

The regularity is a synthesis of observations across domains and does not show that it works in the same proportion everywhere. The size of the difference varies with the field and with the nature of the events compared. What transfers is the direction, not the quantity.

The direction alone yields a practical consequence. The influence of a period of lowered substance outlasts that period. Cut corners for three messages and three will not necessarily restore it.

From here, how frequency reads changes. Raise the frequency and the chance of mixing in a message whose substance cannot be held rises. The mixed-in rounds are recorded more heavily than the good ones. So the cost of raising frequency is not only workload.

For the same reason, “send something anyway” is an expensive judgement. Do not send and the order falls only gradually. Send a round with nothing inside it and it drops a step there. The cost of not sending and the cost of sending a thin round are not the same.

And the difference does not appear in the figures. A month with nothing sent has zero deliveries, so no open rate is calculated. A month with a thin round gets an open rate calculated, which can land near the usual level. On the figures, the month with the thin round looks better.

The influence of a thin round shows up in the figures as the opening of the next round. By then the cause sits in the previous round. The aggregation figures alone give no clue for suspecting the previous round.

The Limits of a Subject Line That Speaks to the Reader’s Difficulty

One widely shared piece of subject-line advice is to touch the reader’s difficulty.

The policy is correct, and its execution carries a condition. The difficulty exists only on the other person’s side.

What can be written here is a name for the difficulty. Where the name matches their state, the subject line functions as their own words. Where it does not, it passes as words directed at somebody you imagined.

And whether the name landed cannot be read from the open-rate figures. Missing, and landing on a day with no room to open, are both recorded as the same “not opened.”

Further, the name of a difficulty has a problem of intensity. Write it strongly and response increases, and by however much it was strengthened, a betrayal follows if the substance cannot answer it.

The betrayal does not appear as a loss in that message. A record of opening remains, so the day’s figures improve. The loss appears as the opening of the next message.

This lands in the same place as the structure treated in product launch, where a reason supplied from outside inflates. A subject line raised in intensity demands at least the same intensity next time.

Then set a restrained subject line? Restrained ones sometimes go unopened.

The line of determination is not intensity but whether the subject line accurately states the content of the body.

An accurate subject line has a secondary effect. It lets the other person correctly make the judgement not to open. Somebody who skipped, knowing today was unrelated, does not lower the name’s position. Somebody who opened and found it unrelated does.

So an accurate subject line does not raise the open rate; it makes the open rate meaningful. The proportion opening approaches the proportion who needed the content.

This effect appears as deterioration in the short-term figures. The better people become at correctly deciding not to open, the lower the open rate goes.

Accurate subject lines are also harder to write. Without a settled account of what the body states, accuracy is impossible. A vague subject line can be written even where the body is unsettled.

So being unable to write the subject line tells you about a problem in the body first. If what was written cannot be said in one sentence, that message is trying to say several things.

Subject-line advice carries a measurement problem too. Opening is often measured by an image load, and how it gets counted varies with the recipient’s settings. Where a pre-loading setting exists, opening is recorded without anybody opening.

Comparing absolute open rates against others therefore carries little meaning. What carries meaning is only the trajectory of your own figure measured under the same conditions. This structure is treated in what the email open rate actually is.

Once a Template Is Identified, It Is Processed Ahead of the Substance

The advice on structure presents a template. Enter from a greeting, place the substance, close with a notice. The template reads easily and genuinely functions.

Used repeatedly, it takes on another function. The recipient starts recognising the template ahead of anything else.

Once recognised, the manner of reading changes. The first paragraph is known to be a greeting, so it goes unread. The last is known to be a notice, so it goes unread. What is substantively read is the middle only.

This is not a failure but normal processing. Against things that arrive repeatedly, people omit steps. Without omission, there is not enough time for the volume arriving.

After the omission begins, what gets observed on your side is a fall in response. The same template and the same volume are being sent, and only the response falls.

The prescriptions often adopted at this point are two: change the template, or raise the frequency.

Change the template and it returns temporarily. Only until the new template is identified. Raise the frequency and the omission strengthens.

There is a third prescription. Build a position within the template that does not get omitted.

A position that does not get omitted is a position known to hold different substance every time. Where the greeting position carries a different specific each time, it is not omitted. Where the notice position carries a different reason each time, nor is that.

So what prevents omission is not novelty of template but unpredictability by position.

Building an unpredictable position increases the labour of writing. The fixed parts can be reused; the parts that differ each time have to be written each time.

And in busy periods, this is the first part to revert to the fixed form. The reversion does not appear in that week’s figures. It appears as a fall in response months later.

If the template gets omitted, better to have no template — without one, the load on the reader rises.

Text where what sits where is unknown each time demands that the reader build the structure before reading. Demand that work and it goes unread.

What is needed is to hold a template and then separate the places where it creates prediction from the places where it does not. Prediction of position may be created; prediction of content had better not be.

Not every position has to be unpredictable. In practice, one suffices.

Where one position differs each time, the message gets opened in order to check that position. Once opened, the other positions enter the field of view too. One reason to open is enough — that is the point of this design.

One Thing per Message — What You Mean to Leave Behind Settles the Length

The advice on structure asks that the message not run too long. The advice is correct, and no basis is supplied.

The basis is not character count. It is what remains inside the reader once they finish.

Where one thing remains, the length needed to explain that one thing is the right length. Where three things remain, nothing may remain.

The reason three get written sits on the material side. On a day with three things to write, three can be written. Narrowing to one requires the judgement to discard two.

The reason they cannot be discarded is plain too. There may be no occasion to write the discarded two. Where the frequency is fixed, the slots for writing are finite.

There is a method for reversing the order of judgement. Decide from what you mean to leave behind rather than from what to write.

Settle on one thing to leave behind and the material becomes only what supports that one thing. What does not support it stays out, however good it is.

This order has one more effect. When the writing is done, what that message did can be said in one sentence. A message where it cannot be said does not get summarised on the other person’s side either.

What does not get summarised does not remain in memory. And what moves the sender’s position in the order is memory.

Narrow to one and the content thins. The opposite happens.

Narrow to one and the volume available for that one thing increases. Line up three and each gets a third.

Thinning happens not when you narrow but when you put everything in without narrowing. All three end up merely touched on.

Not every round can be narrowed to one. Rounds requiring a notice exist.

In rounds where a notice enters, what remains from that round is the notice. Put a notice and substance in the same round and what remains is the notice.

The work of settling on one thing to leave behind has to happen before writing. Try to settle it while writing and you select from the material that could be produced, and the options get constrained by the material.

Settling before writing requires a stock of material. Settle what to leave behind with no stock and the supporting material does not gather, so you end up writing whatever was available that day.

The discipline of one thing per message is therefore supported not by writing time but by the time spent accumulating material. Material arises in daily life and does not arise on the day of writing. In a form where it gets gathered on the day of writing, this discipline cannot be kept. Accumulation has no deadline, so it is also the first thing cut in a busy period. Cutting it produces nothing that week.

Three Things That Change When Replies Are Possible

What gets overlooked in the design of delivery is whether replies are possible.

Many deliveries come from a sender that cannot be replied to. Receiving replies would require an operation for handling them.

Make replies possible and three things change.

First, how you write changes. Where a reply might come back, it becomes harder to write in a one-directional voice. Who you are writing to becomes concrete on the writing side as well.

Second, how the other person receives it changes. Somebody who can be replied to is somebody who might respond. Contact from somebody who might respond sits in a different position from a broadcast.

Third, the information you receive changes. Replies carry information that figures cannot capture. What was unclear, what differed from their situation, what they want to know.

That information is small in quantity. Replies are few compared with the number sent. The contents of that one, though, teach more than the open-rate figure.

And that one is directly usable in what you write next. What was written concretely in a reply may not be that person’s problem alone.

Handling replies costs time; it will not scale once the count grows. There is a scale at which it does not.

Even so, the reply route is sometimes closed before scale becomes the issue. The reason then lies less in the count than in closing making the operation simpler.

Not every reply has to be answered. What cannot be answered can be stated in advance.

A form that cannot be replied to holds no route for receiving information from the other person’s side. Without the route, what remains on your side is only figures for opening and response. Figures teach what happened and not why it happened.

A form that can be replied to constrains how you write. Knowing something may come back, assertions become more careful. By however much they become careful, the momentum of the writing falls.

That constraint looks like a loss and in fact works in the correct direction. Text written without anticipating a reply tends to take a shape where the room for objection has been removed. Text with no room for objection gives the reader nowhere to hold it against their own situation.

Without a place to hold it against, the content does not connect and does not remain in memory. Not remaining, the name’s position does not move. Here the question of replies returns to the question of opening.

Whether to open the reply route looks like an operational judgement and is in fact a judgement about content. From the moment it opens, whether what you wrote landed on anybody becomes knowable. Once knowable, how the rounds that did not land are written changes.

Measure the History Side, Not Today’s Figure

What produces opening cannot be read from the day’s figures.

The first is the difference when you send twice without changing the subject line. Where the contents differ and the form of the subject line is the same, the difference in opening comes from something other than the subject line. Where there is almost no difference, opening is settled by the name’s position rather than the subject line. This determines whether time should be spent improving subject lines at all.

The second is how the first message after a pause gets opened. Where opening does not fall after a gap of some weeks, the name remains inside the order. Where it falls sharply, what was inside the order was not the name but the frequency of arrival.

The third is whether replies contain specifics. “That was helpful” is an impression, not a position. Where replies containing the specifics of somebody’s situation arrive periodically, even in small numbers, the content is connecting to something on their side. Where none arrive, it is being read without connecting.

None of the three looks at the day’s figures. Since opening is a consequence of history, what has to be confirmed is history.

The three carry conditions. All of them require some period and some count. At the stage of a few messages sent, the history itself does not exist.

What can be done at that stage is to confirm, every time, that the subject line accurately states the body. The impression built by the first few messages becomes the origin of the order thereafter.

The first indicator demands sending without changing the subject line; would anybody do that in practice? Sometimes not.

The confirmation requires no special preparation, though. Match the form of the subject line and send everything else as usual.

The reason for not doing it can be less the labour than the awkwardness if a result appears. A result showing almost no difference changes the standing of all the time spent on subject lines so far.

Handing over the number alone is always possible. How many characters to write in, at what interval to send, how much opening counts as enough. Place a number without stating the conditions and that number works as a target from that day. The moment it works as a target, operations satisfying it get searched for. And among the operations that raise opening are ones that merely shrink the list.

Where the hand was put in and where the figure moved are not the same round. As long as that lag exists, judging by the day’s figures tends to select what takes effect the same day. And what takes effect the same day is gathered in the layer with a low ceiling.

Before writing the next message, one thing can be settled. What is the one thing that will remain inside somebody who finishes reading it? If it can be written in one sentence, that message is more likely to remain. If it cannot, it simply has not been settled before writing. What is unsettled does not become settled by finishing.