AI 提示詞： 電子郵件行銷

Generate high-performing email subject lines for: [topic/offer/announcement]

Audience: [who they are, what they care about, relationship with your brand]
Email type: [newsletter / promotional / transactional / re-engagement / cold outreach]
Brand voice: [formal / casual / witty / urgent / conversational]
Current average open rate: [percentage]
My top 3 best-performing subject lines ever: [paste them]
Industry: [your niche]

Generate 30 subject lines in 5 categories:

1. CURIOSITY GAPS (6) — open a loop the reader must click to close
2. SPECIFICITY & NUMBERS (6) — exact figures that feel too specific to ignore
3. PERSONALIZATION (6) — using [first name], company, or behavior triggers
4. URGENCY & SCARCITY (6) — real time pressure, not fake countdown timers
5. PATTERN INTERRUPTS (6) — unexpected formats (questions, lowercase, emoji, one word)

For each subject line:
- Matching preview text (the second line recipients see)
- Character count (flag any over 50 chars for mobile truncation)
- Spam trigger check: flag words that hurt deliverability

Then:
- Rank the top 10 overall with reasoning
- Identify 3 subject lines to A/B test against each other (and why these 3)
- 5 subject line patterns to AVOID in [industry] (with examples of what not to do)

Build a welcome email sequence for new [subscribers / customers / trial users].

Product/service: [what you offer and price point]
How they signed up: [lead magnet / free trial / purchase / webinar / content upgrade]
Main value proposition: [why they signed up — the promise that got them]
Desired end action: [purchase / upgrade / book a call / activate feature]
Sales cycle length: [how long people typically take to convert]
Current welcome sequence: [describe what you have now, if anything]

Design a [5-7] email sequence:

For each email:
1. Send timing (days/hours after trigger) with reasoning
2. Subject line + preview text (2 options each for testing)
3. Full email copy (150-250 words — short, scannable, mobile-first)
4. Single CTA (one action per email, never multiple)
5. Psychological principle driving this email (reciprocity, social proof, loss aversion, etc.)
6. Segmentation trigger: what happens if they click vs. don't click vs. reply

Also include:
- The 'golden window': which email in the sequence is most critical and why
- Deliverability safeguards: plain text ratio, link count limits, from-name strategy
- When to suppress: if they convert mid-sequence, what emails to skip
- Benchmark open/click rates to expect at each stage

Build a cold email campaign for: [what you're selling or what meeting you want]

Target prospect: [title, company size, industry]
What you offer: [your value prop in one sentence]
Why now: [why this prospect needs this now, not later]
Social proof: [clients, results, credentials — your best one-liner]
Sending infrastructure: [email tool, domain age, daily send volume]

Design a 3-email sequence:

1. EMAIL 1: THE OPENER (Day 1)
   - Subject line: short, lowercase, looks like a human wrote it
   - Body: under 80 words, no links, no images, no HTML
   - End with a question, not a calendar link
   - Personalization hooks: what to customize per prospect

2. EMAIL 2: THE FOLLOW-UP (Day 3-4)
   - Subject line: reply to Email 1 thread
   - Body: under 60 words, add one piece of value or proof
   - Still no links — the link comes only if they reply

3. EMAIL 3: THE BREAKUP (Day 7-8)
   - Subject line: creates mild FOMO or closure
   - Body: under 50 words, easy out, no guilt
   - Optional: one link (case study or resource) since this is the last touch

Also include:
- DELIVERABILITY CHECKLIST: spam trigger words to avoid, sending limits, warm-up schedule
- PERSONALIZATION FRAMEWORK: 3 data points to research per prospect (takes 2 min each)
- RESPONSE HANDLING: templates for positive reply, objection, and 'not interested'
- VOLUME MATH: expected reply rates and how many prospects to contact for [X] meetings

Audit my email deliverability and tell me what's hurting inbox placement.

Current setup:
- Email platform: [Mailchimp / ConvertKit / Instantly / Smartlead / etc.]
- Sending domain: [your domain]
- Domain age: [how old]
- Average daily send volume: [how many emails]
- Current open rate: [percentage]
- Current bounce rate: [percentage]
- SPF/DKIM/DMARC status: [configured / not sure / not set up]
- Recent spam complaints: [if known]
- Type of emails: [marketing / cold outreach / transactional / mixed]

Paste a recent email for analysis: [paste full email with subject, body, links]

Audit:

1. AUTHENTICATION CHECK
   - SPF, DKIM, DMARC status and what to fix
   - Custom tracking domain setup
   - Sending IP reputation assessment

2. CONTENT ANALYSIS
   - Spam trigger words and phrases in subject + body
   - HTML-to-text ratio (flag if too HTML-heavy)
   - Link count and link reputation
   - Image-to-text ratio
   - Unsubscribe link presence and compliance

3. LIST HYGIENE
   - Bounce rate assessment (flag if over 2%)
   - Engagement-based segmentation recommendations
   - Re-permission campaign design (for old/cold lists)
   - Sunset policy: when to remove unengaged subscribers

4. SENDING PRACTICES
   - Volume ramp-up schedule for new domains
   - Optimal sending frequency for my list size
   - Warm-up protocol for cold outreach domains
   - Throttling recommendations

5. ACTION PLAN
   - Priority 1 fixes (do today — these are killing deliverability)
   - Priority 2 fixes (do this week)
   - Ongoing monitoring routine (15 min/week)

Design an automated drip campaign for: [lead nurture / onboarding / upsell / event follow-up / re-engagement]

Trigger event: [what starts the sequence]
Audience segment: [who enters this campaign and their awareness level]
Campaign goal: [the specific conversion you want]
Timeline: [total days/weeks the campaign runs]
Existing content assets: [blog posts, case studies, webinars, tools to reference]
Max branches allowed: [2-3 — keep it manageable]

Design the complete campaign:

1. CAMPAIGN FLOWCHART
   - Entry trigger and qualification criteria
   - Each email node with timing
   - Decision branches: opened vs. didn't open, clicked vs. didn't click
   - Exit conditions: when someone leaves the campaign (converted, unsubscribed, or completed)
   - Re-entry rules: can someone go through this campaign twice?

2. EMAIL COPY (for each node)
   - Subject line + preview text
   - Full body copy (150-250 words, scannable)
   - Single CTA per email
   - What content asset to reference or link

3. AWARENESS STAGE MAPPING
   - Which buyer awareness stage each email targets
   - How each email moves the prospect one stage closer to purchase
   - Where to inject social proof, urgency, and objection handling

4. MEASUREMENT
   - KPIs per email (open rate, click rate, reply rate, conversion)
   - Benchmarks to expect
   - When to intervene manually (hot lead signals)
   - Monthly optimization routine: what to test and when to retire underperformers

Design statistically rigorous A/B tests for my email campaigns.

Current metrics:
- List size: [total subscribers]
- Average open rate: [percentage]
- Average click rate: [percentage]
- Average conversion rate: [percentage]
- Send frequency: [how often you email]

Email platform: [what you use]
Biggest performance gap: [where you most want improvement]
Previous tests: [what you've tested before and results]

Build a 4-week testing plan:

1. WEEK 1: HIGHEST-IMPACT TEST
   - What to test and why this moves the needle most
   - Specific hypothesis: 'If we change [X] from [A] to [B], [metric] will improve by [estimated %] because [reasoning]'
   - Control vs. variant with exact copy for both
   - Minimum sample size for statistical significance (95% confidence)
   - How long to run before calling a winner

2. WEEK 2: BUILD ON WEEK 1 RESULTS
   - Test design that leverages what you learned
   - Hypothesis and copy for both variants

3. WEEK 3: DEEPER OPTIMIZATION
   - Test a structural element (send time, segmentation, personalization depth)
   - Hypothesis with expected outcome

4. WEEK 4: ADVANCED TEST
   - Test a behavioral trigger or automation element
   - Hypothesis and measurement plan

Also include:
- TESTING RULES DOCUMENT: when to call a winner, minimum run time, how to handle inconclusive results
- SAMPLE SIZE CALCULATOR: formula for my list size to reach significance
- COMMON MISTAKES: 5 testing errors that invalidate results (with fixes)
- COMPOUNDING MATH: if all 4 tests produce modest wins, what's the cumulative impact?

Help me build (or fix) my newsletter.

Newsletter status: [starting from scratch / existing with problems]
Audience: [who reads this and why they should care]
Topics I cover: [3-5 content areas]
Frequency: [weekly / biweekly / monthly]
Current subscribers: [number, if existing]
Current open rate: [percentage, if existing]
Time I can spend per issue: [hours]
Monetization plan: [sponsorships / premium tier / lead gen / none yet]

Deliver:

1. NEWSLETTER IDENTITY
   - Name options (5) with tagline for each
   - Unique positioning: what makes this different from other newsletters in [niche]
   - One-sentence value prop for the signup page

2. FORMAT DESIGN
   - Repeatable section structure (intro, main story, quick hits, spotlight, CTA)
   - Word count targets per section (keep total under 1000 words for weekly)
   - Visual design recommendations (headers, dividers, formatting)
   - Content sourcing system: where to find material each week in 30 minutes

3. SAMPLE ISSUE
   - Full draft of issue #1 with all sections populated
   - Subject line and preview text

4. GROWTH PLAYBOOK (0 → 1,000 subscribers)
   - 5 free growth tactics ranked by effort-to-result ratio
   - Signup page copy with conversion optimization tips
   - Cross-promotion strategy with complementary newsletters
   - Content upgrade ideas to capture email from blog traffic

5. RETENTION STRATEGY
   - How to maintain 40%+ open rates as you grow
   - Re-engagement triggers for subscribers going cold
   - Feedback collection: how to learn what readers want more/less of

I have [number] subscribers who haven't engaged in [timeframe].

What they originally signed up for: [lead magnet / offer / content]
Last topic they engaged with: [if known from email platform data]
What's changed since they went cold: [new features, offers, content, company milestones]
Current list hygiene: [do you regularly clean your list?]
Platform: [email tool]

Build a 4-email re-engagement sequence:

1. EMAIL 1: THE PERSONAL CHECK-IN (Day 1)
   - From a person's name, not the brand
   - Acknowledge the silence without being guilt-trippy
   - Offer a compelling reason to re-engage (not a discount — value)
   - Under 100 words, plain text feel

2. EMAIL 2: THE VALUE BOMB (Day 4)
   - Best content or resource from the past [timeframe] they missed
   - Curated, not comprehensive — 3 items maximum
   - Each item with a one-sentence hook for why it matters

3. EMAIL 3: THE EXCLUSIVE OFFER (Day 8)
   - Something only for re-engaging subscribers
   - Clear deadline or scarcity (real, not manufactured)
   - Social proof: what other subscribers are getting from staying active

4. EMAIL 4: THE CLEAN BREAK (Day 14)
   - Honest, respectful final email
   - Clear choice: click to stay or do nothing to unsubscribe
   - No guilt, no manipulation — a professional goodbye
   - Explain why you're removing them (deliverability health)

Also include:
- SUNSET POLICY: exact criteria for when to remove unresponsive subscribers
- WIN-BACK METRICS: expected recovery rate (realistic, not optimistic)
- POST-CLEANUP PLAN: what to do with the healthier, smaller list
- PREVENTION STRATEGY: 3 tactics to stop subscribers from going cold in the first place