Email warm-up funnel, email by email

A ready-made setup: a subscriber lands in your list and walks themselves to the purchase. The AI writes the email sequence (intro → value → proof → offer), your email service sends it on schedule, and you watch opens and clicks and tune from there. Build it once and it runs on autopilot.

What you end up with

• A 4–6 email warm-up series with subject lines, preview text and copy, all pointed at one goal.
• An autoresponder: someone leaves their email → the messages go out on their own, day by day.
• A lead magnet and a signup form (with consent — that's the law).
• Metrics: opens, clicks, unsubscribes → you can tell which email is sagging.
• A template you copy for the next product.

What you'll need

Tools from the library (by name): ChatGPT/Claude (email copy, subject lines), the prompt "Warm-up: a 4-email series", the prompt "Cold outreach: B2B first-contact email" (if the funnel is cold/B2B), the prompt "YouTube titles (10 variants for CTR)" (adapt it to subject lines — the subject decides whether it gets opened), n8n (wire the form/payment to your email service and other tools), and the combo "Landing page + payments in one evening" (signup and sales pages). Not in the library but relevant: an email service with autoresponders (any ESP with triggered emails and double opt-in), a signup form on your landing page or in Tilda. Accounts/money/hardware:

• Free route: ChatGPT/Claude (free), an ESP free plan (usually up to N subscribers), Tilda Free for the signup page. Any PC.
• Fast route: a paid ESP plan (more subscribers, segmentation, subject-line A/B tests), a paid model for a more consistent tone, n8n Cloud for the integrations.

Step by step

1. Lead magnet and promise. What to do: decide what someone gets for subscribing (a checklist, a mini-guide, a template) and who needs it. How you know it worked: you're not embarrassed by the lead magnet, it solves a specific pain your audience has and leads naturally to your product.
2. Sequence outline. Tool: the prompt "Warm-up: a 4-email series" in ChatGPT/Claude. What to do: generate the sequence for your goal. Ready-made prompt:

Build a 5-email warm-up series. PRODUCT: "<what I sell>". AUDIENCE: <who, what stage>.
LEAD MAGNET: <what they already got>. GOAL: get them to <purchase/inquiry>.
Logic: #1 intro + quick value (no selling); #2 unpack the pain and how to solve it;
#3 proof/case study/story; #4 offer with a deadline and an objection handled; #5 "last chance".
For each email: Subject · Preview text · Goal · Copy (short, to the point) · CTA · when to send (day).
Keep the value-to-selling ratio tilted toward value. Tone: <which>.

How you know it worked: the emails move from value to offer logically, each has one goal and one CTA, and the deadline at the end is real. 3. Subject lines for opens. Tool: the prompt "YouTube titles (10 variants)" (adapt it to email). What to do: generate 5–10 subject-line options per email — the subject decides the open. How you know it worked: the subject promises a benefit or hooks curiosity, is 50 characters or less, no ALL CAPS and no spam words ("FREE!!!"). 4. Building the autoresponder. Tool: your ESP (autoresponder) plus n8n if needed. What to do: load the emails into the service, set the delays between them (e.g. 0 / +1 / +3 / +5 / +7 days), and hook it to the signup form. How you know it worked: a test subscriber actually receives the emails on schedule, and links and images work. 5. Signup and sales pages. Tool: the combo "Landing page + payments in one evening". What to do: build a landing page with the lead magnet and form, plus a payment page for the finale. How you know it worked: signup works (the lead magnet email arrives immediately) and the payment button actually takes money. 6. Measure and tune. What to do: after a week or two, look at the metrics (open rate, CTR, unsubscribes) per email. Ready-made prompt for the review:

Here are the metrics for my email sequence: <subject — open% — click% for each email>.
Find the weak spots: where opens sag (→ rewrite the subject), where clicks sag (→ rewrite the offer/CTA),
where unsubscribes climb (→ pushing too hard/wrong audience). Give concrete edits for each email.

How you know it worked: you can see which email is the bottleneck and rewrite just that one instead of the whole sequence.

Free route vs fast (paid)

You really can launch the sequence for free on an ESP free plan. Paying starts to pay off as the list grows: subject A/B tests and segmentation lift opens, and n8n ties sending to payments and your CRM. What decides the outcome isn't the budget — it's the warm-up logic itself.

Common problems and fixes

• Emails land in spam. Double opt-in only (confirmed subscription), no ALL CAPS or spam words in subjects, a working unsubscribe link and legal details in every email, and warm up the domain/mailbox gradually.
• Low open rate. The bottleneck is the subject: run 5–10 variants (step 3) and test them; the sender name should be recognizable.
• They open but don't click. Weak offer/CTA — rewrite the body and the call to action (step 6); one CTA per email, not three.
• Unsubscribes climbing. Too many sales emails back to back — tilt the balance back toward value; check that the audience is right (a relevant lead magnet).
• The emails read like spam blasts. Give the AI one or two of your own real emails as a tone sample, write plainly and directly; edit the first line and the P.S. by hand.

Time and money

• Time: building the sequence (copy + subjects + autoresponder + page) — 3–6 hours; tuning against metrics — 20–30 minutes once a week.
• Money: free on an ESP free plan or about $0–30/month for a paid ESP plus optionally a paid model/Tilda/n8n.
• Honestly: the main thing in email is the law (in Russia, 152-FZ and the advertising law): you may only send with consent, with an unsubscribe link and your legal details. A list without consent means fines, not "growth". And the funnel only works with an honest lead magnet and a real audience.