email-marketing

SKILL.md

---
name: Email Marketing
description: Email deliverability, list management, sequences, segmentation, and campaign optimization.
metadata:
  category: marketing
  skills: ["email", "newsletters", "sequences", "deliverability", "automation"]
---

## Deliverability Foundations

- Authenticate all sending domains: SPF, DKIM, and DMARC — missing any one tanks deliverability
- Warm up new domains/IPs: start with 50-100 emails/day, increase 20% daily over 2-4 weeks
- Never buy email lists — purchased lists destroy sender reputation permanently
- Clean list regularly: remove bounces immediately, unengaged after 90 days
- Monitor blacklists: MXToolbox, Google Postmaster Tools — problems compound if not caught early

## List Health Metrics

| Metric | Healthy | Warning | Critical |
|--------|---------|---------|----------|
| Open rate | >20% | 10-20% | <10% |
| Click rate | >2% | 1-2% | <1% |
| Bounce rate | <2% | 2-5% | >5% |
| Unsubscribe | <0.5% | 0.5-1% | >1% |
| Spam complaints | <0.1% | 0.1-0.3% | >0.3% |

- One spam complaint per 1000 emails is the danger zone — above this, inbox placement drops sharply

## Subject Lines

- 40-50 characters optimal — truncates on mobile after 35-40
- Personalization works: name or company in subject increases opens 10-20%
- Curiosity beats clarity for opens, but clarity wins for clicks — balance based on goal
- Avoid spam triggers: ALL CAPS, excessive punctuation!!!, "free", "act now"
- Test emoji: works for some audiences, hurts others — A/B test before committing
- Preview text is second subject line — don't waste it with "View in browser"

## Email Sequences

**Welcome sequence (5-7 emails over 2 weeks):**
1. Immediate: Deliver promised value + set expectations
2. Day 1: Best content or quick win
3. Day 3: Story/origin + values
4. Day 5: Social proof + case study
5. Day 7: Soft offer or deeper engagement
6. Day 10: Address common objection
7. Day 14: Clear CTA

- First email in sequence has highest open rate — make it count
- Space emails 1-3 days apart — daily is too aggressive for most audiences

## Segmentation

- Segment by behavior, not just demographics — what they clicked matters more than job title
- Minimum viable segments: new subscribers, engaged (opened in 30 days), unengaged, customers
- Tag on every click — builds behavioral profile automatically
- Sunset unengaged subscribers after 90 days — send re-engagement, then remove
- VIP segment for highest engagement — reward them with early access, exclusives

## Timing and Frequency

- Tuesday-Thursday mid-morning performs best on average — but test your audience
- B2B: business hours. B2C: evenings and weekends can work
- Frequency depends on value: daily works if valuable, weekly fails if boring
- Let subscribers choose frequency — reduces unsubscribes significantly
- Send time optimization: most ESPs offer this, use it

## Campaign Types

- **Newsletter**: regular value, builds relationship, low direct conversion
- **Promotional**: clear offer, urgency, direct conversion focus
- **Transactional**: receipts, confirmations — highest open rates, add value here
- **Re-engagement**: "We miss you" + incentive + easy unsubscribe
- **Announcement**: product launches, major updates — segment to interested only

## Writing for Email

- One goal per email — multiple CTAs dilute response
- Write for skimmers: bold key phrases, short paragraphs, bullet points
- P.S. lines get read — put secondary CTA or key point there
- Plain text often outperforms HTML for personal-style emails
- Mobile-first: 60%+ open on mobile — single column, large tap targets

## Automation Triggers

- Sign up → Welcome sequence
- Purchase → Onboarding + cross-sell sequence
- Abandoned cart → 3-email recovery (1h, 24h, 72h)
- No open in 30 days → Re-engagement sequence
- Link click → Tag and trigger relevant follow-up
- Date-based → Birthday, renewal reminder, anniversary

## A/B Testing

- Test one variable at a time: subject, send time, CTA, from name
- Need 1000+ recipients per variant for statistical significance
- Subject line tests: 20% of list first, winner to remaining 80%
- Measure what matters: opens for subject, clicks for content, conversions for offers
- Document every test — institutional learning prevents repeat experiments

## Common Mistakes

- Sending to entire list always — segment or face declining engagement
- No double opt-in — leads to fake emails and spam traps
- Ignoring mobile preview — broken layouts kill credibility
- Same email to all segments — personalization is expected now
- Hard selling too early — value first, offer later
- Ignoring unsubscribe feedback — they tell you what's wrong

## Compliance

- Include physical address in every email — required by CAN-SPAM, GDPR
- Unsubscribe must work within 10 days — one-click preferred
- Honor opt-outs immediately — delayed removal is illegal in many jurisdictions
- GDPR: explicit consent required, document it, allow data deletion
- Separate consent for different email types — marketing vs transactional