Favors.dev
A SaaS onboarding email sequence is a short series of automated messages, usually five, that carries a new signup from the account they just created to the first real result your product gives them. Each email has one job: welcome and set expectations, remove the obstacle to the first win, answer the loudest objection, show proof from someone who is not you, and close with an honest check-in. You write it once. It works on every signup after that, including the ones who arrive at two in the morning while you sleep.
Most founders write it too late. The launch goes well, a few hundred people sign up in a week, and almost all of them look at an empty dashboard, decide they will come back to it, and never do. That is not a traffic problem and no amount of additional launch effort fixes it. It is the gap between the account and the outcome, and email is the only channel you own that can reach across it.
Every published guide on this topic shows you screenshots of onboarding emails from companies with a hundred employees and a design system. Those are useful for admiring and useless for copying, because the constraint that shaped them was brand consistency across millions of sends, and your constraint is that you have forty signups and no idea which of them is stuck. What follows is the version that fits that situation: the activation moment you have to define first, five annotated fill-in templates, the trigger logic, the format decision, and the deliverability rules that decide whether any of it lands in an inbox at all.
Why does launch traffic die at the empty dashboard?
Because a signup is a statement of curiosity and nothing more, and curiosity has a short half-life. The person who created an account during your launch was reading about your product thirty seconds earlier. By tomorrow they have read about four others. Whatever motivated them is gone unless something re-creates it, and the only thing you own that can do that is their inbox.
This is the part of launch mechanics that gets the least attention, because it happens after the fun bit. Getting attention is loud and visible and produces a graph you can screenshot. Converting that attention into someone who has actually used the product is quiet, unglamorous work that happens one person at a time. The founders who get compounding results from their first hundred users are almost always the ones who did this part, not the ones who ran a better launch.
There is a second reason to write it now rather than later. The sequence is a forcing function that makes you state, in plain sentences, what your product is for and who it is wrong for. Almost every founder who writes these five emails discovers that two of them are hard to write, and the difficulty is always a positioning problem wearing a copywriting costume.
What is your activation moment, exactly?
Write this down before you write a single email. Your activation moment is the first thing a user does that makes the product obviously worth their time, stated as a specific event your system can detect. Not signup. Not a login. Not a page view. The moment the value lands.
The test is simple. If a user did this one thing and then never opened another email from you, would you still count them as successfully onboarded? If yes, that is your activation moment. If you find yourself saying they would also need to do three more things, you have not picked a moment, you have picked a checklist, and a checklist cannot be the target of a sequence.
Real activation moments
Connected their first data source and saw their own numbers on screen. Published one page. Sent one invoice. Imported a contact list and ran a first search against it. In every case a person saw their own material come out the other side.
Things that only look like one
Completed the profile. Watched the tour. Clicked around the dashboard for four minutes. Invited a teammate. These measure compliance with your interface, not value received, and a sequence aimed at them produces obedient users who churn.
Once the moment is named, the sequence has a spine. Email two exists to remove whatever stands between a person and that moment. Email five exists to find out why it did not happen. Every line you are tempted to add gets measured against it, which is how you avoid the sprawl that turns an onboarding sequence into a newsletter nobody asked for. If you have not yet decided which events you are even capable of detecting, that decision belongs with the rest of your launch metrics, and it is worth an hour before you write copy.
The five emails, as fill-in templates
Fill in every blank with something specific to your product, then read the result out loud. If a sentence sounds like something a company would say rather than something you would say, rewrite it in the words you would use on a call. The annotation under each template is the part most sequences get wrong, so read those before you start editing.
Welcome, and why you built it
The job: Set one expectation and make the next click obvious.
Subject pattern: You are in. Here is the fastest way to ______
Hi ______,
Thanks for signing up for ______. I am ______, and I built this because ______ used to take me ______ and I could never find a tool that just ______.
The fastest way to see whether it works for you is to ______. It takes about ______ minutes and you do not need to ______ first.
[ Button: ______ ]
Over the next week I will send you ______ short emails, and then I will stop. Reply to any of them and it comes straight to me.
______
What it is not: Not a feature tour. Not a list of everything the product does. One reason it exists, one action, one expectation about what arrives next.
The first-win walkthrough
The job: Remove the specific obstacle standing between this person and their first result.
Subject pattern: The ______ step most people get stuck on
Hi ______,
You created an account yesterday but have not ______ yet, which is completely normal. Nearly everyone stalls at the same place: ______.
Here is the shortest version.
1. ______
2. ______
3. ______ (this is the part that does the work)
[ Button: ______ ]
If step ______ is where it broke for you, hit reply and tell me what you saw on screen. I will fix it or walk you through it today.
______
What it is not: Not a nag. It names the obstacle rather than the inactivity, so it reads as help instead of surveillance.
The objection killer
The job: Answer out loud the doubt that is quietly keeping them from committing.
Subject pattern: But what about ______?
Hi ______,
The question I get most often about ______ is ______.
The honest answer: ______. What we do instead is ______, and here is what that looks like in practice: ______.
If ______ is your situation, this is probably not the right tool yet, and I would rather tell you that now. If it is not, the thing to try next is ______.
[ Link: ______ ]
______
What it is not: Not a sales page. The admission of who the product is wrong for is the part that makes the rest believable.
Proof from someone who is not you
The job: Let a third party say the thing you cannot say about yourself.
Subject pattern: How ______ uses this to ______
Hi ______,
______ runs ______ and had the same problem you probably signed up with: ______.
What they do now: ______. It took them about ______ to set up, and the part they mentioned unprompted was ______.
“______” ______, ______
If that is close to your setup, the equivalent for you is ______.
[ Button: ______ ]
______
What it is not: Not a logo wall and not a five-star quote with no name attached. One named person, one specific job, one detail nobody would invent.
The honest check-in
The job: Convert the actives, learn from the inactives, and close the sequence cleanly.
Subject pattern: Last one from me (did ______ work?)
Hi ______,
This is the last automatic email you will get from me.
If ______ is working for you, the next thing worth doing is ______.
If it is not, I would genuinely like to know why. Reply with one word and I will read it: too slow, too confusing, missing ______, or wrong fit.
That reply is worth more to me than the signup was.
______
What it is not: Not a discount. Not a fake deadline. The one-word reply mechanic is there because it is the only feedback request a busy person will actually answer.
The blanks are doing more work than they appear to. Every one of them is a place where a generic sequence would have inserted a sentence that sounds fine and means nothing. Filling in the blank in email three, in particular, is uncomfortable on purpose: naming who your product is wrong for is the single fastest way to become believable to everybody else.
When should each onboarding email send?
Behavior beats the calendar wherever you can detect behavior. A purely time-based sequence sends the walkthrough email to someone who already finished the walkthrough, which is a small insult that costs you the next three opens. One conditional rule, on the activation event you defined above, fixes most of it.
| Fires | Condition | Branch | |
|---|---|---|---|
| 01 Welcome | On signup, within seconds | Everyone | Never skip this one |
| 02 First-win walkthrough | 24 hours later | Only if the activation event has not fired | Activated users skip straight to email 03 |
| 03 Objection killer | Day 3 | Everyone still subscribed | Send to activated users too, the doubt survives activation |
| 04 Proof | Day 5 | Everyone still subscribed | Swap the example to match their signup segment if you have one |
| 05 Honest check-in | Day 8 | Everyone still subscribed | Different final paragraph for activated and stalled users |
Send in the recipient's working hours, not yours. A solo founder in Bangkok emailing buyers in Chicago at a convenient local time is sending every message at three in the morning their time, where it lands under twenty overnight newsletters. If your tool can stagger by timezone, turn that on before you touch anything else in the schedule.
Plain text or designed?
Plain text, for almost everything you send at this stage. The advantage is not aesthetic and it is not about open rates. It is that a plain message from a named human invites a reply, and at forty users every reply is worth more than the email that produced it. A designed template with a header image and a footer full of social icons quietly tells the reader that no one is home.
| Factor | Plain founder mail | Designed template |
|---|---|---|
| Who it looks like it came from | A person who runs the company | A marketing department |
| Reply rate | People answer, because answering is obviously possible | Replies feel like shouting at a billboard |
| Time to write and change | Minutes, in any tool, including your own outbox | A template build, then a rebuild every time copy changes |
| Rendering risk | Almost none, it reads the same everywhere | Dark mode, image blocking and older clients all break it |
| Where designed wins | Poorly, once you need a real screenshot or a product visual | Anything visual: a UI walkthrough, a chart, a changelog |
There is one exception worth taking seriously. When the content is genuinely visual, a screenshot of the exact screen a user is stuck on will outperform three paragraphs describing it. Embed the image in an otherwise plain message rather than reaching for a template system. You get the clarity without turning the email into a broadcast.
Send from a real address at your own domain, with your own name in the from field, and make the reply-to an inbox you read. The no-reply address is the single most common unforced error in founder-run onboarding, because it discards the only feedback channel that reaches people who are too polite to complain in public and too busy to fill in a survey.
What makes an onboarding email actually arrive?
Authentication, a clean list, and a working unsubscribe. This is the layer that vendor guides skip because their platform handles it, and the layer that bites a founder who wired up their own sending on a fresh domain the week of launch.
Start with the records. Google's email sender guidelines are direct about it: set up SPF, DKIM and DMARC for your sending domain, and expect messages that are not authenticated to be treated as suspicious. This is a one-time DNS job that takes under an hour and is the difference between a sequence that reaches inboxes and one that quietly fills spam folders while your dashboard reports successful sends.
Then the unsubscribe. Modern mailbox providers expect a bulk sender to support one-click unsubscribe, which is a real technical standard rather than a link in your footer: RFC 8058 defines the headers that let a mail client remove someone from your list without them ever opening the message. Any reputable sending platform sets these for you. If you built your own sending, they are your job.
The legal layer is smaller than founders fear and not zero. The Federal Trade Commission's CAN-SPAM compliance guide turns on the primary purpose of a message. Purely transactional or relationship content, such as confirming an account someone just created, is exempt from most of the requirements, while any message whose primary purpose is promoting your product must carry a valid physical postal address and honor opt-outs within ten business days. Onboarding emails routinely mix both, so put an unsubscribe link and an address in all five and stop thinking about it.
One habit protects all of this: never import a list. The sequence is for people who typed their own address into your signup form minutes ago. Adding anyone else raises your complaint rate, and complaint rate is the metric that decides whether your domain gets to keep reaching inboxes at all.
Test the sequence before your users do
Onboarding email has a cruel property: the people best qualified to tell you it is broken are the ones who already left. By the time the data says email two is not working, a hundred signups have gone through the version that did not work. The only way around that is to have people read it before it ships, and the only readers worth anything are ones who have written a sequence themselves.
You cannot self-edit this. You know what your activation moment is, so you cannot read email two as a person who does not. A founder who has never seen your product reads the same five emails and tells you that step two assumes a thing you never explained, that email three answers an objection nobody has, and that email five sounds like it was written by a tool. That is the feedback that changes outcomes, and it is real work to produce, which is exactly why strangers do not hand it out for free.
That asymmetry is the reason Favors.dev exists. It is a founder marketing co-op running on a points economy: you earn points by doing verified marketing favors for other founders, including structured feedback on assets exactly like a five-email sequence, and you spend those points requesting the same back. Because the feedback action is graded and attached to a real profile, what comes back is a specific critique of email three rather than a thumbs up. Post the request in the favors queue and it goes in front of people who have shipped their own.
The same loop covers the stage before this one. If you are still collecting addresses ahead of a launch, the sequence those people receive is a different animal, and the waitlist marketing playbook covers what to send between the signup and the door opening. The five emails here start the moment the door opens.
One more reason to get the replies flowing. The answers people send to email five are where testimonials, review copy and case-study quotes come from, and those artifacts do double duty now that answer engines assemble their responses from attributed third-party material rather than your own landing page. Analyses of chatbot citations consistently find that the sources engines lean on are community posts, review platforms and named-author pages, which are precisely the things a reply from a happy user lets you create.
Frequently asked questions
How many emails should a SaaS onboarding sequence have?
Five is the working number for an early-stage product, and three is defensible if your product has one obvious action. The count matters far less than what each email is for. A sequence fails when two emails share a job, because the second one has nothing new to say and trains the reader to stop opening. Give each email exactly one job: welcome and set expectations, remove the obstacle to the first win, answer the loudest objection, show proof from someone else, then close the loop and ask what happened. If you cannot name the distinct job of a sixth email, you do not need a sixth email.
When should you send the first onboarding email?
Within seconds of signup, while the person is still on the page. The welcome email is the only one in the sequence with a guaranteed audience, because the reader just typed your name into a signup form and is watching for confirmation. Delaying it by even an hour spends that attention for nothing. Everything after the welcome should be triggered by behavior rather than the calendar: the walkthrough email only makes sense for someone who has not activated yet, and sending it to a person who already succeeded reads as proof you are not paying attention.
Should onboarding emails be plain text or designed HTML?
Plain text, for almost everything an early-stage founder sends. A plain message from a founder gets replies, and replies are the entire point of onboarding email when you have a few dozen users, because each one tells you where the product breaks. Designed templates earn their place when the content is genuinely visual, like a screenshot walkthrough or a chart. The mistake is reaching for a designed template because it feels more professional. At your size, looking like a marketing department is a downgrade, not an upgrade.
Do onboarding emails need an unsubscribe link?
In practice, yes, and the safe rule is to include one in every message in the sequence. United States law turns on the primary purpose of the message: an email whose content is purely transactional or relationship, such as confirming an account a person just created, is exempt from most requirements, while anything whose primary purpose is promoting the product is a commercial message and must carry a valid physical postal address and a working opt-out. Most onboarding emails mix both, which puts them in the judgment zone. An unsubscribe link costs you nothing and removes the question entirely.
What is the difference between onboarding emails and a drip campaign?
An onboarding sequence serves one specific outcome, which is getting a new signup to their first real result in your product. A drip campaign is any timed series of emails, including ones with no outcome attached beyond staying in touch. The practical difference shows up in how you decide what to send: every onboarding email is judged against whether it moves someone closer to the activation moment you defined, so an email that is interesting but irrelevant to that moment gets cut. A drip campaign has no such filter, which is why drips sprawl and onboarding sequences should not.