What Should the First Five Emails After Signup Actually Say?

What should the first five emails after signup actually say?

One: confirm and get them to the first action. Two: help them past the step most people stall on. Three: show one real outcome someone like them achieved. Four: ask a question a human will answer. Five: make the decision easy before the window closes. Five emails, five jobs, no overlap.

Most onboarding sequences I audit are five emails that all do the same thing, which is explain features in descending order of how proud the team is of them. That is a product tour split into instalments, not a sequence.

A sequence should move someone from signed up to succeeding. Here is how I structure it, and the technical requirements that decide whether any of it arrives at all.

Why do most onboarding sequences fail?

Because they are written from the inside out. The team knows what the product does and writes emails explaining that. The user does not yet know what they are trying to achieve with it, so feature explanations land as homework rather than help.

The second reason is that each email is written independently. Somebody drafts five emails in an afternoon and each one restates the value proposition because each one was written as if it might be the only one read. That is defensible logic and it produces five versions of the same email.

The third reason is that nobody defined the outcome. If you cannot say what a successful sequence looks like in terms of a user action, you cannot tell whether it is working, so it never gets improved and sits there for three years.

What should email one do?

Confirm they are in and point at exactly one next action. Not a tour, not a list of everything available, one action with a link. The single most valuable thing email one can do is remove the paralysis of an empty account by deciding for them what to do first.

Make the action the smallest possible step toward value, not the biggest. If the real payoff needs a data import, do not open with import your data. Open with something they can finish in two minutes that makes the next step feel reasonable.

Send it immediately and make it look like it came from a person. A transactional-looking welcome with a logo header and four navigation links reads as automated, and automated email trains people not to open the next one.

What should emails two and three do?

Email two removes the specific obstacle your data says people get stuck on. Email three shows one concrete outcome from a customer who resembles them. Together they handle the two reasons people stall: they cannot get past a step, or they are not convinced it is worth the effort.

For email two, you need to know the stall point, which means looking at where signups stop rather than guessing. There is almost always one step with a visible drop. That step is the entire subject of email two, and it should read like help rather than a nudge.

Email three is proof, and the mistake is making it a case study. Nobody reads a case study in an email. One sentence about who they are, one sentence about what changed, one link if they want more. The resemblance matters more than the size of the result, which is the same principle behind content that helps your champion sell internally.

What should emails four and five do?

Email four asks a genuine question and is written so a reply is the obvious response. Email five names the decision point and makes it easy to act on. Four is where you learn things. Five is where you convert the people who were already close.

Email four is the one almost nobody sends and it is my favourite. A short, plain-text message from a real person asking what they were hoping to do with the product. The replies are the best research you will get, because they come from people mid-evaluation rather than from a survey months later.

Email five acknowledges where they are. If the trial ends soon, say so and say what happens next. If they have used the product properly, reference that. If they have not, offer a different path rather than pretending. How you frame that depends partly on how long your trial runs and what happens at expiry.

How do you time these?

Trigger on behaviour where you can and fall back to a schedule where you cannot. A sequence that sends email two about a step the user already completed is worse than sending nothing, because it proves you are not paying attention.

The minimum viable version is branching on one condition: did they complete the first action. Those two paths need genuinely different emails, not the same email with a different opening line. Even that single branch outperforms a straight five-email drip in my experience.

Whether you run this in Customer.io, HubSpot, Intercom, Loops, Mailchimp or something you built yourself matters less than whether the tool can branch on a product event at all. If it cannot see what the user did inside the product, you are writing a newsletter with a signup trigger.

On spacing, I stay close early and widen later. The first days after signup are when intent is highest and the account is still novel. Waiting three days to send the second email hands that attention to whatever else they signed up for that week.

What technical requirements decide whether these arrive?

Authentication, reputation, and unsubscribe handling. Google's sender requirements say all senders must set up SPF or DKIM for their sending domains, have valid forward and reverse DNS records, use a TLS connection, format messages per RFC 5322, and keep spam rates in Postmaster Tools below 0.3 percent.

If you send more than five thousand messages a day, Google classifies you as a bulk sender and the bar rises. Bulk senders must set up both SPF and DKIM, must set up DMARC for the sending domain, and must ensure the domain in the From header is aligned with either the SPF domain or the DKIM domain.

Google also requires that marketing messages and subscribed messages support one-click unsubscribe and include a clearly visible unsubscribe link. That requirement applies to your lifecycle emails, not only to your newsletter, and it is the one teams most often assume does not apply to them.

None of this is optional and none of it is your email tool's job to do for you. Your sending domain, your DNS records and your From header are yours. The spam rate threshold is the number I would put on a dashboard. It is measured in Postmaster Tools, which means you need to have set that up before you need it. Discovering your spam rate during a deliverability crisis is a bad way to meet the metric, and none of your copy matters if the message lands in spam.

How do you know the sequence is working?

Measure the action, not the email. Opens and clicks tell you about subject lines. What you want to know is whether the share of new signups who complete the key action within their first week went up after you changed the sequence.

That means your product analytics and your email tool need to agree on what happened. Whether you use Amplitude, Mixpanel, PostHog or a warehouse, the activation event has to be defined once and referenced everywhere, or you will end up with two numbers and an argument. Pick one activation event and hold it steady. Changing the definition halfway through makes every historical comparison meaningless, and it is a surprisingly common way teams convince themselves something worked.

Reply rate on email four deserves its own place on the dashboard, even though it looks small. A sequence generating real replies is generating conversations, and conversations at this stage are worth more than a marginally better click rate on email three.

And watch what happens to the people who never activate. Those are the ones your sequence failed, and they are usually the larger group. What you learn from them is more actionable than another optimisation on the path that already works, which is the same argument as running a proper win-back campaign for churned customers.

What should you do next?

Open your current sequence and write the single job of each email in one sentence. If two emails share a job, one goes. If any email's job is explain our features, rewrite it around the user's next action instead. That exercise alone fixes most sequences.

Then check your authentication and your Postmaster Tools spam rate before you touch the copy. The best onboarding email in the world does nothing from a spam folder, and this is the part that is objectively either correct or not.

If you want someone to look at your sequence and tell you honestly which emails are earning their place, reach out. It is usually a one-conversation problem with a two-week fix. Let's chat.