Onboarding email sequence for SaaS: one task per email
Build an onboarding email sequence backwards from activation: one task per email, sent while the task is open, stopped at activation, judged by clicks.
An onboarding email sequence is 4 to 6 emails, each asking for one task, sent while that task is open and stopped at activation. Picture a founder on Monday: the welcome email shows a high open rate, but most trials never build a first dashboard. The email looks healthy, and the product data says it is not doing its job.
What an onboarding email sequence is for
An onboarding email sequence has one job: move a new user to the activation event. The activation event is the first product action that shows the user got real value. Every email in the sequence should push toward that action or build on it.
A newsletter works differently. It sends the same content to the whole list on a calendar. An onboarding email reacts to what one user did, or did not do, since sign-up.
Transactional mail is different again. Password resets and receipts confirm something the user just did. An onboarding email asks the user to do something next.
The common "tips" drip fails because it ignores behavior. Everyone gets a feature tour on day 2, 4 and 6. The user who already activated reads about a step they finished. The stuck user reads about a feature they cannot reach yet. Nobody measures the drip, because no product event is attached to it.
Open your current sequence this week. Next to each email, write the product event it should cause. Mark every email where you cannot name one.
Start from the activation event, then work backwards
Write the activation event down before you write a single subject line. Make it a tracked event in your analytics, with a name and a clear rule. "Built a first dashboard on their own data" works. "Got familiar with the product" does not.
Take a hypothetical B2B analytics tool with a free trial. Its activation event is the first dashboard built on the user's own data. Two tasks sit in front of it: connect a data source, then build the dashboard. One step stalls most users: the data import.
Now list the tasks between sign-up and activation. Most products have two or three. Each task becomes one email, and the step where users stall gets its own unblock email. In the example, two tasks plus the unblock email give three emails before activation.
Activation tasks live in the product, so the email can only be as good as the screen it points to. The SaaS onboarding best practices guide covers the in-product side: empty states, checklists and the first session.
Write your activation event and the tasks before it on one page this week. Circle the task where your funnel data shows the biggest drop.
Send on behavior, not on the calendar
Every email needs two rules: a send condition and an exit condition. The send condition says the email goes out only if its task is still open. The exit condition says the user leaves the pre-activation emails the moment the activation event fires.
Timing becomes a delay, not a trigger. "Day 3" means three days after sign-up, if the task is still open. A user who connected data in the first hour never gets the import help on day 3.
Lifecycle tools such as Customer.io or HubSpot handle this with an entry trigger, a check before each send and an exit rule. The setup only works if your product events reach the email tool. Without those events, every send falls back to the calendar.
Activated users move to a second track. Emails 4 and 5 build the habit, then bring in the team or a plan. A user who never activates gets emails 1 to 3, then leaves the sequence. Sending them upgrade asks would only teach them to ignore you.
Write a one-line spec for each email this week: the task, the send condition and the exit condition. Check that every event in that spec reaches your email tool.
A five-email onboarding sequence example you can adapt
Here is the full sequence for the hypothetical analytics trial. Days count from sign-up. Emails 1 to 3 go only to users who have not activated. Emails 4 and 5 go only to activated users.
| Timing | Send condition | Subject line example | The one ask | Link target | |
|---|---|---|---|---|---|
| 1. Welcome | Minutes after sign-up | Not activated, no data source connected | Connect your data to see your first dashboard | Connect a data source | Data connection screen |
| 2. First result | Day 1 | Data imported, no dashboard yet | Your first dashboard is one template away | Build a dashboard from a template | Template picker, data already selected |
| 3. Unblock | Day 3 | Import started, not finished | Stuck on the import? Here is the fix | Retry the import with the guide, or reply | Import screen with the help panel open |
| 4. Habit | Day 5 | Activated | Get this dashboard in your inbox every Monday | Schedule a weekly report | Report schedule settings |
| 5. Team or plan | Day 7 | Activated | Share your dashboard with the people who need it | Solo users: invite a teammate. Teams: pick a plan | Invite screen, or the plan page |
Email 5 still carries one ask per person. The send condition picks which ask a user sees, so nobody gets both.
The upgrade ask waits on purpose. A user who has not built a dashboard has no reason to pay for one. Asking anyway trains them to skip your emails, including the one that could have unblocked them.
Late activators need one extra rule. Send email 4 on day 5 or one day after activation, whichever comes later. Send email 5 two days after email 4.
Copy this table into a doc today. Replace each task, subject line and link target with your own, and keep the send conditions as they are.
How to write each email
Each email gets one ask and one button. The button deep-links to the exact screen where the task happens, with the user logged in. A link to the dashboard home makes the user hunt for the task, and many will not.
Send from a plain sender name: a person at your company, not the brand alone. Set the reply-to to an inbox someone reads. Replies to the unblock email are free user research, so answer them within a day.
Keep the body short. Two or three sentences above the button is enough. Explain why the task matters to the user, not to you.
Write the subject line as the task, not the brand. "Connect your data to see your first dashboard" tells the reader what happens next. "Welcome to Acme" tells them nothing they did not already know.
Here is email 1 from the table, written out:
Subject: Connect your data to see your first dashboard
Hi Sam, your workspace is ready. One step is left: connect the data source you want to report on. Once it is in, your first dashboard is one template away.
[Connect a data source]
Ana, product lead. Reply to this email if the import gives you trouble.
Rewrite your welcome email this week to one ask, one deep-linked button and a sender who answers replies.
Measure clicks and product events, not opens
Open rates do not tell you whether an onboarding email worked. Apple's guide says Mail Privacy Protection "prevents senders from seeing if you've opened the email message they sent you." The same page notes that the protection "does not extend to links." Clicks remain a usable signal, and product events are the one that counts.
Vendor benchmarks show how far opens and clicks sit apart. In Mailchimp's benchmarks, all users average a 35.63% open rate and a 2.62% click rate, with data last updated in December 2023. Mailchimp counted only campaigns sent to at least 1,000 subscribers, and it warns that Apple's Mail Privacy Protection may affect open rate accuracy.
Mailchimp's figures cover marketing campaigns across industries, not onboarding emails. Use them to see the gap between opens and clicks, never as a target for your sequence.
Read each email with two numbers. The first is its click rate, compared with the other emails in your own sequence. The second is the share of recipients who complete the email's task in the product within 48 hours.
The two numbers point to different fixes. Few clicks means the email is the problem: subject, timing, the ask or a buried button. Many clicks with few completed tasks means the product step is the problem. The link lands on a wall, and better copy will not move it.
The bigger lift often sits in the product flow, not in the messages. For Masa, a consumer app, Kreto restructured the flow to surface the core experience first and defer registration. The case page names that change as the primary driver of sign-up conversion rising from 32.7% to 63%. The same page describes a welcome sequence in OneSignal that introduced key features progressively after sign-up. That is the same one-thing-at-a-time pattern this post recommends.
When the diagnosis points at the product step, Kreto's product marketing work fixes the email and the screen it lands on as one funnel.
Pull both numbers for every email in your sequence this week. Put them in one table and mark each row: fix the email, fix the product step or keep it.
Get to the inbox first
None of the measurement matters if the email lands in spam. Gmail and Yahoo both publish sender rules, and they differ in scope.
Google's email sender guidelines say that starting February 1, 2024, all senders to Gmail accounts must meet its requirements for all senders. Two of them: set up SPF or DKIM for your sending domains, and keep spam rates reported in Postmaster Tools below 0.3%.
Google adds a volume rule. Senders of more than 5,000 messages a day to Gmail must support one-click unsubscribe for marketing and subscribed messages. They must also include a clearly visible unsubscribe link in the message body.
Yahoo's Sender Best Practices list similar requirements for all senders: SPF or DKIM at a minimum, and a spam rate below 0.3%. For bulk senders, Yahoo asks for a list-unsubscribe header with one-click unsubscribe for marketing and subscribed messages. Yahoo also asks bulk senders to honor unsubscribes within 2 days.
The two unsubscribe rules have different scopes. Google ties its rule to a daily volume to Gmail. Yahoo frames its rule around senders it calls bulk. Your sending volume decides whether either rule applies to you, so add one-click unsubscribe to onboarding emails anyway. An easy exit beats a spam report.
Yahoo also asks senders to "set recipient expectations clearly when users subscribe." For onboarding, the sign-up screen is that moment. One line under the form, saying a few setup emails will arrive in the first week, does the job.
Check your SPF and DKIM records and your Postmaster Tools spam rate this week, before you change any copy.
Common mistakes
Most broken sequences fail in the same few ways.
- Calendar-only sends. Activated users get "finish your setup" on day 3. Add a send condition to every email.
- Upgrade asks before value. The trial user who never built anything gets a pricing email. Keep the plan ask for activated users.
- Several calls to action. Three buttons split the click and blur the reading. Keep one ask per email.
- Sending from no-reply. The stuck user cannot reply, so you lose the most useful feedback you get. Use a person's name and a reply-to someone reads.
- Judging by opens. A high open rate hides an email that moves nobody. Read clicks and task completion.
- Never retiring an email. An email that fails both numbers for a month still goes out. Rewrite it or delete it.
Pick the mistake your sequence shows most and fix it before your next send.
FAQ
How many emails should an onboarding sequence have?
Four to six works for most trial products, but the count follows your tasks. Write one email per task before activation, one unblock email for the step where users stall, and one or two after activation. If a task does not exist, the email should not either.
When should the first email go out?
Send it within minutes of sign-up, while the user is still near the product. Send it only if setup is still open. Make its one ask the first setup task, with a button that opens that exact screen.
Should onboarding emails have an unsubscribe link?
Yes. Google requires one-click unsubscribe for marketing and subscribed messages from senders of more than 5,000 messages a day to Gmail. Below that volume, add it anyway. A user who can leave easily does not need the spam button.
What is a good click rate for onboarding emails?
The useful comparison is inside your own sequence. Rank your emails by click rate and look hard at the lowest one. Then check task completion within 48 hours, because a click that ends on a broken screen is no win.
Map your sequence to the activation event
An onboarding email earns its place when it moves a user to a product event. Define activation, give each task one email, send while the task is open and stop at activation. Then read every email by clicks and completed tasks.
Start this week with your activation event and the tasks before it. If you want a second view on mapping your sequence to that event and reading it per email, talk to Kreto.