Email
AI Prompt for Curiosity Subject Lines
This AI prompt for email subject lines writes curiosity-driven options that earn opens. Get 15 lines that tease value without clickbait tricks.
The prompt
Role: You are an email copywriter obsessed with open rates. You write subject lines that create an information gap the reader must close by opening, never with clickbait or deception. Context: The email is about [EMAIL_TOPIC] and is being sent to [AUDIENCE]. The one thing the reader will learn or get inside is [INSIDE_VALUE]. The tone of my brand is [BRAND_TONE]. Task: Generate 15 curiosity-driven subject lines. Each one should tease the value inside without giving it away, using techniques like open loops, surprising specifics, and questions. Vary the structures: some questions, some statements, some with numbers, some without. Format: A numbered list of 15 subject lines. After the list, mark your top 3 picks and give a one-line reason for each pick. Constraints: Every subject line under 50 characters so it shows fully on mobile. No clickbait: the email must deliver what the subject promises. No ALL CAPS, no excessive punctuation, no spam trigger words like 'free', 'guarantee', or 'act now'. No emojis unless [EMOJI_OK] says yes. Each line must make sense to someone who has never heard of my brand.
Variables to fill in
Replace each [variable] in the prompt with your own details:
[EMAIL_TOPIC] [AUDIENCE] [INSIDE_VALUE] [BRAND_TONE] [EMOJI_OK]
Example input
EMAIL_TOPIC: our guide to cutting software costs; AUDIENCE: startup founders; INSIDE_VALUE: 7 negotiation scripts that saved companies $50k+; BRAND_TONE: friendly and direct
How to use this prompt
Run this every time before you send an important email, then A/B test your top 2 picks on a small sample before sending to the full list.
Tips for the best results
- Under 50 characters keeps the full line visible on phones.
- Specific numbers in subject lines lift open rates.
- Always deliver on the curiosity or trust erodes fast.
Frequently asked questions
How many subject lines should I test?
Test 2 at a time on 10 to 20 percent of your list each, then send the winner to the rest. Testing more than 2 splits your data too thin.
Do questions or statements open better?
It depends on your audience, which is why you test. Questions often win for educational content, statements for announcements.