How to Write a Launch Email That Gets Opened

You spend three months building the thing and eleven minutes writing the launch email that tells people it exists.

A launch email has one job: earn one click from somebody who already knows who you are. Every paragraph either serves that click or gets cut before it sends.

That sounds easy until you count the people. A list of 1,400 that took two years to build delivers to roughly 1,340 after bounces and filtering, and a normal campaign clicks at 2.62% (Mailchimp's benchmark table, December 2023 data, read 24 October 2026). That is 35 people.

So the work happens before the writing: choosing who gets the email, timing it against their week, and finding the one line they act on. What follows is the list cut, the five-line body, the send sequence, and the three numbers to check at 48 hours.

Four Gates Between the Draft and the Click

Four gates stand between your draft and the click, and any of them can end the send before a word of copy is read.

The spam folder comes first, and it is decided by your sending history rather than the words in your subject. Then the tab: Gmail files promotional mail into Promotions, which is a real inbox but a quieter one. Then the subject line, cut at about 33 characters on a phone. Then the first two lines, which decide whether anyone scrolls.

Most advice about launch email behaves as if only the last two gates exist. That is why so much of it spends its length on subject lines and closes with "add a clear call to action", and why a well-written email can still send 24 people to a page that was never built for them.

The failure that shows up most often is a launch email that is really three emails stapled together: the announcement, the feature list, and the founder's thank-you note. It contains no single ask, so it earns no single click.

Try to write this one sentence before you open the editor.

After reading this email, [which people] will [do exactly one thing] at [where].

Fill all three brackets or the email is not ready. Once you can, the draft has a spine, and every paragraph that does not point at those brackets comes out. On a 1,400-person list that pruning usually takes the body from 400 words to 150, and the click rate rises, because there are fewer places to stop reading.

The Four Decisions, in the Order They Cost You

Four decisions run a launch email, and they are rarely the ones anyone spends an afternoon on.

Run them in that order, because each one constrains the next. A subject written before the segment is picked is a guess about who is reading. A page picked before the ask is decided is a homepage.

Do it in one sitting, ninety minutes, a week before the product is finished.

1. Build the segment in your email tool and write its size down. Nine minutes, and that count decides everything after it.
2. Find the hour. Open your last ten campaigns, sort by click rate, and note the hour each one sent. Four minutes. No history at all means 7am in the reader's time zone.
3. Write the subject to 40 characters, then cut it again to 33. If the meaning dies at 33, the subject was two ideas.
4. Build the page: one headline, the price or the date, one button, no navigation.
5. Paste the three counts and the hour into a note you will still have on launch day.

That ninety minutes is the difference between an email written by someone who thought about it and one written at midnight because the build finished late.

Cut the List Before You Write a Word

The launch is the one send a small brand cannot afford to lose, and the way it gets lost is volume.

Validity's 2025 Email Deliverability Benchmark Report puts global inbox placement at 87.2%, with 6.1% landing in spam and 6.6% going missing, and North America slightly worse at 86.7%. Microsoft's consumer mailboxes are the hardest room in the house: 75.6% inbox.

Google's sender guidelines ask bulk senders to keep the user-reported spam rate under 0.3%, and have required one-click unsubscribe in commercial mail since 1 June 2024. Cross those lines and the next campaign pays for it, not this one.

So cut the list first, then write.

1. Build a segment called Core: opened or clicked in the last 90 days.
2. Build Warm: opened or clicked between 91 and 365 days ago.
3. Treat everything older as Cold and keep the launch email away from it entirely.
4. Remove hard bounces and anyone who has ever complained, permanently, not just from this send.
5. Write the three counts down beside the hour from the last section.

Here is the arithmetic on a list of 4,300 built over four years: 980 Core, 210 Warm, 3,110 Cold. Send the launch to Core only. Of 980 addresses around 2% bounce or get filtered, so call it 960 delivered. At 2.62% that is 25 clicks. Send the same email to all 4,300 and you deliver 4,214, maybe earn 40 clicks, and collect complaints from people who last heard from you in 2023. At the 0.3% ceiling that is 12 complaints, enough to spend the next month in the spam folder.

Twenty-five clicks that arrive beat forty that cost you the quarter.

And the 3,110? They get nothing this month. They get one plain re-permission email in a few weeks, asking whether they still want the occasional note, with no launch news in it at all.

The instinct to fight is the announcement reflex: the product is live, so everyone must be told. Most of those people are not readers any more. If the Cold pile is more than 40% of your list, the launch goes to Core and the real work this month is building a list from zero again, behind an offer worth an address.

The Subject Line, With the Truncation Math

Two numbers govern the subject line and neither of them is about cleverness.

Aim for 30 to 50 characters, with the words that carry the meaning inside the first 33, because that is where a phone cuts it. Long subjects win in tests run on desktop screens and lose in the inbox people actually hold.

The second number is the one almost nobody accounts for. Apple's Mail Privacy Protection, live since September 2021, preloads images on Apple's servers, which fires the tracking pixel whether or not a human opened anything. Mailchimp's documentation on it says opens are unreliable for those contacts and lets you exclude them from reports from 22 June 2024 onward. Clicks were never touched.

Judge the subject by clicks, then. Opens will flatter you for as long as you let them.

Three shapes do most of the work, and all three are short:

• The reader's problem in their words: "Your invoices are three days late".
• A number they can act on: "Two seats, $8 a month, live now".
• A plain sentence from a friend: "I built the thing I kept complaining about".

Then test it in a way that gives you an answer instead of an opinion.

1. Write five subjects, not one.
2. Cut each to 40 characters, then push the meaning into the front 33.
3. Send all five to yourself and read them on a phone, in your own inbox, next to your bank's alerts.
4. If the list is over 1,000, send the best two to a 10% sample, wait four hours, and pick the one with more clicks.

The tricks to skip are the ones that borrow trust they never earned: a fake "Re:", a deliberate typo, an emoji pretending to be a person. They work once, and the second send is the one that matters.

How to Write a Launch Email in Five Lines

The body is 120 to 180 words. Longer than that and it stops being an email and becomes a page nobody asked for.

Five lines do it.

Subject: Two seats, $8 a month, live now

Preview: The one thing that changed, in a sentence.

Line one: the news, stated inside the first six words. Line two: who it is for, and what it replaces. Line three: the link, written as a sentence ("The page with the price is here"). Line four: what happens after the click, with the number attached. Line five: the ask, or your name and nothing else.

Plain text, from a person's name rather than the brand's, with the reply address set to an inbox somebody reads. One link, and the same link if you repeat it anywhere.

Below 500 subscribers, swap the button for a request: ask them to reply with the one thing that would make the product worth paying for. A launch to 200 people that produces nine replies tells you more than 200 clicks, and it costs nothing to set up.

Send one to a friend before you send it to the list. If they cannot say what you are asking them to do after a single read, the email has four ideas in it and needs to lose three.

Where the Click Sends Them

The email is worth exactly what the page behind it is worth.

Four rules for that page, and every one of them removes a decision:

1. The headline repeats the promise of the subject in a fuller sentence. Same words, more of them.
2. The price or the date sits above the fold, in text rather than in an image.
3. One button, repeated once at the bottom, with the same words both times.
4. No navigation. No logo link to the homepage. Nowhere else to go.

The failure is always the same: a launch email that sends 25 careful clicks to a homepage, where the reader has to find the new thing among the old things. They will not. Send them to a page with one job, and if that page does not exist yet, spend the ninety minutes there instead of on the copy. Conversion rate optimization with low traffic is the arithmetic of why that page decides more than the subject line does.

A launch email reaches the people who already know you, and a launch usually needs one more place where a stranger can find the thing. That place is part of the same week's work. One of them is ours: The Board is a leaderboard of brands and a paid placement for every brand on it, with rank sorted strictly by what each has paid and by nothing else. 25 brands hold a row right now, each one a mark, a line of copy and a link out, and the higher the amount, the more impressions the row earns. It sits beside the email in the plan rather than in place of it: the email reaches the people who opted in, and the board puts the same brand in front of the ones who never will.

Send Times, the Second Send, and Who Gets It

Launch day has one send in it. The launch has five, and the second one is where most of the clicks come from.

The hour is your hour, not the internet's. Every guide has a best-time-to-send chart built from somebody else's list. Yours is in your own reports: the sends with the highest click rate, read against the hour they went out. With no history at all, 7am in the reader's time zone is a reasonable guess, because it puts the email above the day's noise rather than inside it.

Then the second send, at 48 hours, to the people who did not click.

Send it to the non-openers instead and it lands on people who read the first email and chose not to act, because Apple's preloading makes an unopened list partly fiction. Clicks are the honest signal, and they came through the privacy changes intact.

Two thresholds keep the resend from becoming a habit that hurts:

• Above 1% clicks on delivered, send the second, with new information: what the first 48 hours produced, a customer's sentence, a real deadline.
• Under 0.5%, send nothing. The subject did not fail, the offer or the page did, and a second send doubles the complaints without doubling the clicks.

Six Sends, in the Order They Land

One audience each, one link each. This is the whole campaign, and it fits into a morning's writing if the product is done.

1. Seven days out, Core only. The news and the date, no price, one link to a page that takes an address for a reminder. This is the email that makes launch day work, and it is the one people skip.
2. The day before, Core and Warm. The hour it goes live and what it will cost, with the same reminder page.
3. Launch day, at your hour. The five-line email, one link.
4. Two days later, non-clickers only. New proof rather than a reminder: the first ten customers, the thing that broke, the number you hit.
5. Five days later, Core and Warm. What the week produced, in one honest figure, with the same link. People buy when something is visibly working.
6. Day seven, only if there is a real deadline. The price change or the cohort closing, stated plainly. Invented urgency is a debt the next launch pays.

A one-day launch with no deadline drops the last two and runs four sends. A launch with a waitlist adds nothing: the same six, with the waitlist page swapped in until day zero.

If the sequencing is the part you keep rebuilding, the startup launch checklist has the dates in the order they need to happen, call to action examples has the button lines for when you cannot face writing another one, and the rest of our launch posts are on the same shelf.

The Scorecard: Three Numbers at 48 Hours

Forty-eight hours is the reading, and three numbers are enough. The rest of the report is vanity wearing a suit.

Take the 1,340-person list from the top. At 2% clicks, 27 people arrive. If the page converts at 8%, a modest rate for traffic that clicked on purpose, two of them buy. At $49 a month that is $98 of recurring revenue from one email.

Which is the number that reframes the exercise. Two customers is not a launch, and it is what a single email buys on a small list. The second and third sends are the campaign rather than a courtesy. The list you cut and the page you built are what make each one worth sending, and the copy is the smallest part of the whole thing.

Write the sentence. Cut the list. Send it twice to the right people, and read three numbers on Thursday.