Lifecycle Email Sequences

You build the automated emails that move people from one stage to the next: subscriber to customer, signup to active user, buyer to repeat buyer. Good sequences react to what the person does, stop the moment their job is done, and are proven by comparison with a group that did not receive them. You design and draft; switching a flow on is an external action the owner approves.

When to use

• "Set up a welcome series."
• "We need onboarding emails for new trial users."
• "Write a nurture sequence for leads who download the guide."
• "Recover abandoned carts."
• "Win back customers who stopped buying."
• "Our automated emails are old. Review them."

What you need

• The email or marketing automation tool and its triggers: connected app (Mailchimp, HubSpot) or the owner's signed-in browser. Check which events and properties it can see (signup date, purchases, product usage, plan, last activity).
• The lifecycle stages and the action that marks success at each (first purchase, first project created, upgrade, second order). Ask product and sales with ask_teammate.
• Data on where people drop off: product analytics (PostHog, Mixpanel, Google Analytics) or the store backend.
• Positioning, voice guide, customer stories and objections, to fill emails with real substance.
• Consent and legal basis per country; which emails are transactional (order and account messages) and which are marketing (need consent and an unsubscribe). Confirm the rules for the owner's countries with web_search.
• If the email tool is neither connected nor signed in, design and draft the flows as documents for the owner to build, and ask with clarify which events and properties the tool can see.

Method

1. Map the lifecycle. List the stages (subscriber, lead, trial, new customer, active, at risk, lapsed). For each, write the success action and where people get stuck. Pick the one or two flows with the biggest gap between people entering and people succeeding; build those first.
2. Define each flow precisely with the flow card in references/sequence-blueprints.md:
   • Goal: the single action that ends the flow successfully.
   • Trigger: the event that enters someone (signed up, started trial, left cart with value over X).
   • Entry filters and suppressions: who is excluded (existing customers, people in another flow, unsubscribed, recent purchasers).
   • Exit conditions: goal reached, unsubscribed, entered a higher-priority flow.
   • Timing and send windows in the recipient's time zone.
   • Branches: what changes if they did or did not do the step the last email asked for.
   • Frequency cap across all flows (for example no more than one automated marketing email per day per person).
3. Give every email one job. One purpose, one primary CTA. Order the jobs along the path to the goal: deliver what was promised, get a quick first win, show how others succeeded, remove the main objection, make the offer.
4. Write from evidence. Use the customers' own words for the problem, real stories and real numbers (with permission), and the objections sales hears most. Each email should be useful even if the reader does not click.
5. Use behaviour, not only time. Send the next email because of what the person did or did not do ("created a project" vs "has not created a project after 2 days"), with time-based fallbacks. Behaviour-triggered emails usually beat fixed calendars because they arrive when they are relevant.
6. Set up a holdout. Where volume allows, keep a random 5 to 10 percent of entrants out of the flow (or out of the new version) so you can measure the real effect on the goal. Without it, you only see who converted, not who converted because of the emails.
7. Build and test in the tool. Draft the flow in the email tool with every email, delay, branch and exit; keep it off or paused. Run test contacts through every branch; check tokens, links, UTMs, unsubscribe, and that exits fire. Check that transactional emails are not marketing in disguise.
8. Get approval to switch on. Show the owner the flow map, the emails and the test results. Turn it on only after an explicit go-ahead, and decide whether existing contacts enter or only new ones.
9. Measure per flow and per email. Primary: goal conversion rate versus holdout, and time to goal. Per email: click rate, unsubscribe rate, complaint rate, and the step's conversion. Review weekly for the first month with a cronjob, then monthly. Fix the email with the biggest drop-off first.

Output

• Lifecycle map with stages, success actions and drop-off data.
• For each flow: flow card, a diagram (text or mermaid) of emails, delays, branches and exits.
• Each email: subject options, preview text, body, CTA, send condition, UTM content tag.
• Test log showing each branch run with a test contact.
• Measurement plan with holdout size and review dates. Blueprints and timings to adapt are in references/sequence-blueprints.md.

Checks before you finish

• Every flow has one goal, a trigger, suppressions and exit conditions.
• People who reach the goal leave the flow at once.
• No one can get more than the agreed number of automated emails per day across flows.
• Every email has one job and one primary CTA.
• Marketing emails have consent and an unsubscribe; transactional emails stay transactional.
• Branches were tested with test contacts.
• A holdout is in place, or the reason for not having one is written down.
• The flow is off until the owner approves it.

Pitfalls

• Calendar-only drips. Sending day 1, 3, 5 regardless of behaviour wastes the best signal. Branch on actions.
• No exits. Sending "start your trial" to someone who already bought annoys them and raises complaints.
• Flow collisions. A person in three flows at once gets three emails a day. Set priorities and a frequency cap.
• Selling in every email. Earn the ask with useful emails first; then make a clear offer.
• Measuring without a holdout. Flows take credit for conversions that would have happened anyway.
• Set and forget. Product changes break links and screenshots. Review every flow at least quarterly.
• Discount reflexes. Leading cart and win-back flows with discounts trains people to wait for them. Start with help and reminders, test an incentive later.