Email Newsletter and Lifecycle Playbook

One list and one sending setup can do two jobs: a newsletter that keeps future buyers close until they are ready, and lifecycle sequences that help customers get value from what they bought. Map every email a person can get from you, keep a consent record for every address, send from one authenticated setup, start each sequence on something the person did and end it on an exit, cap what anyone gets in a week, and judge each stream by replies, pipeline and retention rather than opens.

Startup email grows one sequence at a time. The newsletter sits in one tool, the checklist follow-up in another, onboarding emails go out by hand, and nobody adds up what a single person receives. In my view, most email problems at seed are not bad copy; they are streams that don’t know about each other.

This playbook treats the newsletter and the lifecycle sequences as one system on one list. It is for people who opted in or already pay you: cold outreach to strangers runs on its own domain, under its own rules, and stays out of this setup; the Cold Outbound Playbook covers it. Writing a good issue is covered in How to Grow a Startup Newsletter, and texts and WhatsApp in SMS and WhatsApp Marketing for Startups.

Start with an inventory, not a new sequence. List every automated email and every recurring send: the newsletter, each sequence, product updates, receipts, password resets and anything sales sends from a template. For each one, note who gets it, what triggers it, how many emails it holds, who owns it and when someone last read it.

Then sort each stream by its job. A stream either builds an audience before it buys, which is the newsletter’s job, or helps a customer get value, which is the job of the lifecycle sequences; a stream that does neither is a candidate to switch off. Mark transactional email separately: it doesn’t count toward the marketing load, but it lands in the same inbox.

Lay the streams against the stages a person moves through: reader, lead, in a deal, new customer, active customer and lapsed (Figure 1). Two problems show up at once: stages with no sequence, and stages where several fire together. In my view, a complete map holds six sequences beside the newsletter: welcome, onboarding, nurture, expansion, win-back and sunset.

Keep it all on one list: one record per person, one consent record and one suppression list that every stream checks. The Lifecycle Email Audit is built for this inventory, and its coverage sheet names the next gap to fill.

One list serving two jobs needs one place where permission lives. Give every address five fields: where it came from, the date, the words the person agreed to, the basis you rely on, and their country. Every stream checks that record, and the suppression list, before it sends (Figure 2).

The FTC sorts email into commercial content and transactional or relationship content, and the primary purpose of a message decides which it is^[1]. Receipts, password resets and notices about changes to a service the customer pays for are relationship messages. An onboarding email whose subject line sells an upgrade is commercial, and needs everything a marketing email needs.

Existing customers can often be marketed to without a fresh opt-in, within limits. In the UK, the ICO’s soft opt-in covers someone who bought or negotiated to buy a similar product from you, provided they had a clear chance to opt out when you collected their details and get one in every message^[2]; Article 13(2) of the ePrivacy Directive sets a similar exception for the EU^[3]. In Canada, an existing business relationship implies consent for two years after a purchase, and for a subscription the two years start when it ends^[4].

In the US, CAN-SPAM doesn’t require consent before a commercial email, but Gmail’s guidelines say not to send to people who didn’t sign up^[5]. My policy for founders is simple: opt-in for the newsletter everywhere, and the customer exceptions only for messages about the product the customer already uses.

Honor opt-outs fast. Yahoo asks bulk senders to honor unsubscribes within 2 days^[6], and Google calls 48 hours a reasonable timeline^[7]. The legal maximum is 10 business days under both CAN-SPAM and CASL, and the opt-out link must keep working for 30 days after a send under CAN-SPAM and 60 under CASL^[1],[4]. In my view, an unsubscribe should take effect the moment it is clicked, automatically.

Clean the list as you go. Yahoo asks senders to monitor hard and soft bounces and inactive recipients, and to remove invalid addresses promptly^[6]; I remove an address at its first hard bounce, and send inactive readers to the sunset sequence in Step 5. This is orientation, not legal advice.

The newsletter and the lifecycle sequences share a reputation, so give them one setup and look after it. Send them from one email platform connected to your CRM, on a subdomain of your main domain such as news.example.com, and send product email such as receipts and password resets from another. In my view, separate subdomains make a problem in one stream easier to see and contain; cold outreach belongs on a separate domain altogether.

Know which rules apply to you. Gmail asks every sender for SPF or DKIM, valid forward and reverse DNS, TLS and a spam rate below 0.3%, and senders of 5,000 or more messages a day to personal Gmail accounts also need DMARC, a From domain aligned with SPF or DKIM, and one-click unsubscribe on marketing and subscribed messages^[5]. Yahoo’s bulk-sender rules match: SPF and DKIM, a DMARC policy of at least p=none, one-click unsubscribe and a spam rate below 0.3%^[6].

Two details catch founders out. Google counts mail from subdomains toward the primary domain’s 5,000, a sender that crosses the line once is treated as a bulk sender permanently, and Google began rejecting non-compliant mail in November 2025^[7]. So set up all three records and the unsubscribe headers before you need them (Figure 3).

Then watch the spam rate. Google asks senders to keep the rate reported in Postmaster Tools below 0.10% and never reach 0.30%^[5]. The Email Authentication Checker reads your SPF, DKIM and DMARC records and shows what is missing.

The newsletter is the one stream a person joins before they want anything from you, so it is where most of your list starts. The reader, the promise and the shape of each issue are in the newsletter article; this step is about the newsletter’s place in the system.

Give people a reason to sign up today, not only a promise about future issues. The best sign-up offer is useful the moment they confirm: a checklist, a template, or a short email course like AAJ’s own ten-lesson course, sent over three weeks. Tie it to the newsletter’s promise, so the people it attracts want the issues too.

Then send a short welcome, two or three emails at most, that delivers the offer, says what arrives and when, and asks one question a reader can answer in a line. In my view, replies are the most useful thing a newsletter produces at seed: each one tells you what to write next, and some come from buyers.

Pick a cadence you can keep, and record where every subscriber came from in a hidden field on each form. Source is how you judge growth: a channel that brings fewer readers who reply and ask for demos is worth more than one that brings many who never do (Figure 4). A reply that mentions buying goes straight to sales, inside the response times in the Lead Management Playbook.

Write each sequence as a short spec before any copy: the trigger, who it is for, each email with one job and one link, the exit, and the number it should move. Start sequences on something the person did, such as signing a contract, finishing a step or going quiet, rather than on a date. Every sequence needs an exit, and every email in it should skip itself once its job is done.

Build in the order the gaps cost most. For a sales-led startup that is usually onboarding first, because a new customer who doesn’t reach value is the most expensive person to lose, then nurture, expansion and win-back. The Customer Onboarding & Retention Playbook defines the first value moment the onboarding emails point at, and AAJ’s free Email Lifecycle Sequence skill, listed under Use with this playbook, drafts the missing sequences.

Nurture runs on the reason a lead wasn’t ready: one short track per reason code, the newsletter underneath, and a return to sales the moment the lead’s score crosses your threshold. Expansion runs on usage, comes from a named person, and pauses while a support ticket is open; an upgrade offer to someone who is stuck is worse than none.

Win-back opens with one question and no offer, because the reason someone left decides whether anything you send will matter. Sunset protects the list: readers who haven’t clicked, replied or logged in for 180 days get one email asking whether they still want yours, and leave if they don’t answer. That window is my rule of thumb, not a benchmark.

Each sequence looks reasonable on its own; the trouble is the person who qualifies for three at once. Before a new sequence goes live, follow one real person through their first weeks and count every marketing email they would get from every stream. Show the transactional ones beside them, because they share the inbox, but don’t count them.

Set a cap and the rules that keep you under it. My rule of thumb, not a benchmark: no more than three marketing emails a week to one person. Then set the priority: one sequence at a time per person, onboarding before everything, a lead in an open deal gets the newsletter only, and a sequence that would break the cap waits.

Delay a waiting sequence rather than drop it, and recount each time you add one. The Lifecycle Email Audit’s send-load sheet does the adding up per audience.

Each stream gets the measure that matches its job. For the newsletter, that is replies, demo requests from subscribers and the pipeline they start; for onboarding, the share of new customers who reach first value on time; for nurture, recycled leads that return to sales; for expansion and win-back, the accounts that grow or come back. Over all of them sits list health: spam complaints, bounces and unsubscribes per send.

Leave open rates off the page. Apple says Mail Privacy Protection downloads remote content in the background by default, rather than only when you open an email^[8], so a tracking pixel can load for a message nobody read. In my view, an open rate that mixes people and machines can’t tell you whether a stream is working.

Test new sequences against doing nothing. My rule of thumb, not a benchmark: hold a random tenth of the people who qualify for a new nurture or win-back sequence out of it for a quarter, and compare the two groups. At startup volumes the difference is often too small to be sure of, so read it as a direction and change one thing at a time.

Review every stream in one 30-minute meeting a month, on one page (Figure 7). End the meeting with one change, write it down, and check it at the next review.

What changes by stage?

What mistakes should I avoid?

• Adding sequences without a shared map. Each team’s stream is reasonable on its own; the person in three of them isn’t (Step 1).
• Adding contacts nobody asked. Webinar registrants and customer users need their own yes before marketing reaches them (Step 2).
• Sending marketing from the domain you use for cold outreach. One bad month can reach your newsletter and your receipts (Step 3).
• Judging the newsletter by subscriber count. Count replies and demo requests by source (Step 4).
• Sequences that run on dates and have no exit. A customer who already finished a step still gets the email about it (Step 5).
• Launching a sequence without counting the load. Six emails in a first week read as noise, whatever each one says (Step 6).
• Reporting open rates. They mix people and machines (Step 7).