How many emails should a nurture sequence have?

Three to seven, set by the buyer's questions, not a rule.

What actually sets the number of emails?

A nurture sequence is a short set of automated emails that follows one action a person took, such as downloading a checklist or starting a trial, and moves them toward one next step. The number of emails is not a style choice. It is set by two things: how considered the decision is, and how many real questions the buyer has to settle before they act.

A considered decision involves more people, more money and more risk if it goes wrong. A buyer comparing two accounting tools for a five-person team has maybe three questions. A buyer replacing a security platform across a company has a dozen, and most of them are about other people: legal, procurement, the team that has to use it. More questions means more emails, spaced further apart.

The rule that falls out of this is simple. Count the questions. Give each email one. Stop when the person acts. If you cannot name the question an email answers, cut the email.

How many emails for each kind of nurture?

The table below covers the four cases we see most often. The email counts are planning ranges, not results. The "stop rule" column is the action that should pull a person out of the sequence the moment it happens.

Notice that the demo request gets fewer emails than the trial even though the decision is bigger. That is because a demo request already has a person attached. The sequence's job is to make sure the demo happens and lands well, not to carry the whole sale. For the enterprise case, the sequence is doing a different job: staying useful across weeks while the buyer works through other people.

How do you write one question per email?

Start from the buyer, not from your product. Write down the questions a real person in that situation would ask before saying yes, in the order they ask them. Then assign one to each email. Here is the list for someone who downloaded a checklist on onboarding new customers:

1. Email 1, same day: "Did I get what I asked for?" Deliver the checklist, point to the one section most people skip, and say nothing else. No pitch.
2. Email 2, day 3: "Does this work for a company like mine?" One short example of the checklist applied to a real situation, with the tradeoff it exposed. A link to a related article is fine. A link to a demo is too early.
3. Email 3, day 7: "What would it take to do this properly?" Lay out the options honestly: do it yourself with the checklist, hire for it, or work with a partner. Offer one clear next step and end the sequence.

If you find yourself writing email four for this case, ask which question it answers. Usually it answers none; it repeats email three with more urgency. That is the email people unsubscribe on.

How many days between nurture emails?

Space emails by how long the decision takes, not by how eager you are. A trial that ends in 14 days can take an email every one to three days because the reader is actively inside the product. A buyer running an evaluation across a quarter needs a week or two between emails, or the sequence finishes before the buyer has even had the internal meeting.

Two practical rules. First, the gap should be long enough for the reader to do what the last email suggested. If email two asks them to try a template, give them days, not hours. Second, the gaps can widen as the sequence goes on. Early emails arrive while interest is fresh; later ones should feel like a check-in, not a chase. Your platform makes this easy to set: Kit's delay step holds a subscriber for minutes, hours, days or weeks, and Mailchimp's time delay step does the same job.

When should the sequence stop?

The sequence stops when the person does the thing it was built to cause. That is the stop rule, and it should be set in the platform, not kept in someone's head. The common names for it are goals and exits. In HubSpot, a contact who meets a workflow goal is pulled out before the next action runs, and anyone who already meets the goal when they would enrol is never enrolled at all. In Kit, an exit step removes the subscriber from the automation, and you can branch on events like a tag being added or a purchase. In Mailchimp, an exit condition is an event or action that removes a contact from a flow.

The point is the same in every tool: define the win, and make the win end the sequence. The usual wins are a reply, a booked meeting, a purchase or upgrade, a form submitted, or a tag set by a salesperson. If your stop rule is "they stop opening," it is not a stop rule, for a reason covered in the measurement section below.

Sequences also stop for reasons you do not choose. Every email must carry a clear way to opt out, and under CAN-SPAM an opt-out has to be honored within 10 business days. If you send more than 5,000 messages a day to Gmail, Google expects one-click unsubscribe and a spam rate held below 0.30%, with 0.10% as the target. A sequence that is too long for the decision is one of the fastest ways to earn complaints, so the stop rule protects your sending as well as your reader.

When is one email enough?

One email is enough when the action the person took already answers their question. Someone who downloaded a pricing comparison does not need five emails about why pricing matters. Send the file, offer one next step, and let your regular newsletter do the slow work of staying in touch. One email is also enough when you have no second question worth answering. Padding a sequence to look thorough is how it ends up feeling automated.

The exception is the first contact with a new subscriber, which has its own job. A welcome sequence sets expectations and earns the second open; it is not a nurture sequence, even though the two are often confused.

When should the sequence hand to a person?

Hand over as soon as a human signal appears: a reply, a question in the trial, a forwarded email, a visit to a pricing or security page. From that moment, an automated email arriving in the same thread reads as careless, and a slow human reply reads as disinterest. The stop rule and the handover are the same event, which is why the person who answers replies has to be named before the sequence goes live.

For a demo request, the handover should happen at the request itself. The sequence confirms, prepares and reminds; a person runs the demo and everything after. For an enterprise evaluation, the sequence can keep sending useful material for weeks while a salesperson works the account, as long as the salesperson can pause it with one click when a live conversation starts. This is how nurture fits into the wider set of B2B email campaigns: it carries interest until a person can. A launch sequence works the other way round, with a date doing the job the stop rule does here.

One sequence Sawa built in 2026, for a fractional CFO firm's weekly newsletter for founders: five welcome emails, routed by a four-question survey on the thank-you page, with a default path for anyone who skips the survey. The first draft branched further than that. The client's review showed a list her size could not fill the branches, so the final version has fewer paths and a first email that asks for one thing.

What should you measure?

Measure actions, not opens. Apple's Mail Privacy Protection prevents senders from seeing whether an email was opened, and other clients preload images, so an open rate on a nurture sequence tells you very little about whether a specific person is moving. The numbers that matter are the ones tied to the stop rule:

• Replies, including the ones that say "not now," which tell you how long "now" really is for this audience.
• Bookings and meetings held, not just links clicked.
• Exits by goal: how many people left the sequence because they acted, and at which email. HubSpot reports this as a goal conversion rate, counted only for contacts who were sent an email in the workflow.
• Unsubscribes and complaints by email, because a spike at email four tells you the sequence is one email too long for this decision.

Read the exits by email number. If most people act after email two, the sequence may be a third longer than it needs to be. If almost nobody acts until the last email, the earlier ones are not answering real questions. If people go quiet entirely, that is a job for a separate re-engagement sequence months later, not for more nurture emails now.

What about "seven touches"?

You will see confident claims that buyers need some fixed number of contacts before they buy. We have not found an original study behind the version that circulates, so we do not repeat it. The honest version is that the number depends on the decision, and that the person deciding does not count your emails. They count their unanswered questions. Answer those, one at a time, and stop.

Need help with this? See how Sawa handles nurture and campaign emails, or book an intro call.