AI Prompts for Email Marketing Campaigns That Get Opened

Before: The Weak Prompt

Most email marketing AI guides are wrong about what makes a good prompt. They tell you to add words like "engaging" and "personalized" to your prompt. Those words do nothing. The AI doesn't know what engaging means to your specific audience, and it genuinely doesn't know what personalized means without actual data.

Here's a prompt seen in at least a dozen "AI email marketing" tutorials:

Write an engaging email marketing campaign for a fitness brand.
Make it personalized and use a friendly tone.
Include a subject line and email body.

This is what AI prompts for email marketing look like when they're built around vibes instead of structure.

Why It Fails

The prompt above fails on four specific counts.

No audience signal. "Fitness brand" describes a $5 billion industry with wildly different customer types. A 22-year-old athlete and a 55-year-old trying to lose weight after their doctor's warning are both "fitness brand customers." They need completely different emails.

No campaign context. Is this a welcome email? A promotional push? A re-engagement sequence? Each type has different length norms, CTA placement, and psychological goals. An AI without this context defaults to a promotional email that reads like a coupon flyer.

No subject line constraints. Mobile email clients cut subject lines at 35–40 characters. "Engaging" subject lines in AI output often run 60+ characters, meaning the crucial end of the line gets truncated on every phone.

No deliverability awareness. Spam filters flag specific trigger words. The AI will happily write subject lines containing "FREE!!!" and "100% guaranteed" — neither of which would survive a spam filter — unless you tell it not to.

⚠️ Common mistake: Treating email copy as just "writing" without understanding that deliverability is a copywriting problem. Words in subject lines affect inbox placement. Subject line prompt specifications aren't optional.

After: The Improved Prompt

You are an email copywriter specializing in [brand type] audiences.

Campaign context:
- Email type: [welcome / promotional / re-engagement / nurture / post-purchase]
- List segment: [describe who's receiving this — behavior, purchase history, or profile]
- Goal: [click to product page / redeem offer / book call / read article]
- Offer or hook: [what makes this email worth opening today]

Write this email:

Subject line (write 5 options):
- Maximum 40 characters each
- No spam trigger words (free, guarantee, %, winner, urgent)
- At least 2 options that lead with curiosity, 2 that lead with benefit
- Label each with character count

Preview text (write 3 options):
- Maximum 85 characters
- Should complement the subject line, not repeat it
- Creates an information gap that forces the open

Email body:
- Opening line: Never start with "I hope this email finds you well" or the brand name
- Paragraphs: Maximum 3 sentences each
- One CTA, placed after the main value has been communicated
- CTA text: Specific action ("See the full collection" not "Click here")
- Total length: [short: 100–150 words / medium: 200–300 words / long: 400+ words]

Tone reference: [paste 2–3 sentences from your best-performing previous email, or describe: "direct and dry" / "warm and casual" / "urgent and no-nonsense"]

What this does: Gives the AI a complete creative brief — identical to what you'd give a human copywriter — so the output is immediately usable rather than a starting point for a 45-minute rewrite session.

⚡ Pro tip: The preview text field is the most underused lever in email marketing. Most marketers leave it auto-populated with the first line of the email body. A well-crafted preview text that creates an open loop ("And the one thing I almost didn't include...") can lift open rate by 6–12% on its own. Write it separately, always.

Breaking Down Each Element

Subject line generation: Five options is not too many. Subject lines have the highest variance of any email element — which one wins an A/B test is often surprising. Give yourself options.

The spam trigger constraint is actually a better writing constraint than "be engaging." Avoiding spam words forces simpler, more direct language that tends to be more compelling anyway.

Preview text as an independent element is a less-discussed insight: in Gmail, Outlook mobile, and Apple Mail, the preview text appears before the recipient opens the email. It's a second subject line. Treating it as filler wastes roughly 40% of your "first impression" real estate.

The "never start with brand name" rule addresses a real pattern. Emails that open with "Hey, it's [Brand]!" train readers that the first line contains no value — which trains them to skim-delete.

⚡ Pro tip: Run a separate prompt for email sequences rather than generating each email individually:

Write a 5-email welcome sequence for [product/audience].
Email 1: Deliver the lead magnet, set expectations for what comes next
Email 2 (day 2): One specific insight that reframes how they think about [problem]
Email 3 (day 4): Social proof — one customer story, 150 words max
Email 4 (day 6): Address the top objection to purchasing
Email 5 (day 9): Make the offer with a specific deadline

Each email: subject line options (5), preview text (2), body copy (medium length).
Maintain consistent voice across all 5. Reference previous emails briefly to create continuity.

Variations for Different Contexts

Re-engagement emails (inactive subscribers): Open with acknowledging the gap. "You haven't heard from us in a while" outperforms pretending nothing changed. Prompt: "Write a re-engagement email that acknowledges the subscriber has been quiet. Don't guilt them. Give them a reason to come back that has nothing to do with our product — give them value first."

Post-purchase emails: The highest-trust moment in customer communication. Prompt: "Write a post-purchase email for someone who just bought [product]. Skip the receipt recap. Write as if they're now a member of something, not a transaction that's been processed."

Save and Reuse This

Email prompts with your brand voice reference baked in become dramatically more valuable over time. As you generate more emails and identify which tones and structures outperform, you refine the prompt — and the next campaign starts from a better baseline.

Store your finalized email prompt templates in PromptABCD so anyone on your team can produce on-brand email copy without needing to know how to prompt from scratch.

⚡ Pro tip: Build a 'swipe file' of your best-performing subject lines with open rate data attached and feed it into your email prompt as 'VOICE EXAMPLES — these subject lines achieved [X%] open rate with [audience].' The AI pattern-matches to your actual performance data rather than generic best practices.

Writing Full Email Sequences with AI Prompts

Single email prompts are useful. Sequence-level prompts are where AI actually changes your email marketing operation.

Write a 7-email post-signup nurture sequence for [product/audience].

Subscriber profile: [who just signed up + what they signed up for]
Sequence goal: [trial conversion / purchase / discovery call]
Timeline: Emails 1, 2, 4, 7, 10, 14, 21 days post-signup

For each email:
- Subject line (5 options, max 40 chars, no spam triggers)
- Preview text (max 85 chars)
- Opening hook (first sentence only — make it specific to their situation)
- Core body (max 200 words)
- CTA

Content arc:
- Emails 1–2: Deliver value, build trust, no selling
- Emails 3–4: Introduce the problem they signed up because of
- Emails 5–6: Position our solution with evidence
- Email 7: Make the offer

Voice consistency note: Maintain [tone description] throughout. Reference earlier emails briefly in later ones to create narrative continuity.

What this does: Creates a narrative through-line across 7 emails instead of 7 disconnected messages — which is what separates sequences with 22%+ trial conversion from sequences that produce 8%.

⚡ Pro tip: The "reference earlier emails briefly" instruction is the single most impactful addition to sequence prompts. Callbacks like "Last week I mentioned X — here's why that actually matters for Y" create continuity that makes subscribers feel like they're reading a story, not receiving a drip campaign. Engagement rates reliably improve when emails feel connected.

Store your sequence prompt structure in PromptABCD. A well-built sequence prompt, tuned to your brand and audience, is genuinely reusable across product lines and audience segments with minimal modification.

Prompt for Transactional and Behavioral Email Triggers

Trigger-based emails — sent in response to a specific action like abandoning a cart, completing a milestone, or downgrading a subscription — are the highest-converting email type in most businesses. They're also the hardest to write because they need to feel timely without feeling surveilled.

Write a triggered email for a subscriber who [specific action: abandoned cart / hit a usage milestone / hasn't logged in for 30 days / just downgraded].

Trigger context: What the subscriber did, when, and what it means behaviorally
Tone constraint: This email must feel like it came from a human who noticed the action, not an automated system that logged it
Do not: Reference the technical trigger directly ("We noticed you abandoned your cart")
Instead: Reference the implied state ("It sounds like you're still deciding — that's totally fine")

Write 2 versions:
Version 1: Curiosity-led (create interest in completing the action)
Version 2: Value-led (give something useful that makes completing the action feel natural)

Subject lines: 5 options each, max 40 characters

What this does: Addresses the biggest reason triggered emails feel creepy — they reference the technical event instead of the human feeling behind it. This reframe consistently improves open rates on behavioral triggers.