# 5,000-Contact Email Blast Tripwire — Playbook

**Tier scope:** Tripwire #1 — `tripwire-5k-email-blast` ($247)
**Time budget:** 2-3 person-hours total (offshore) + ~30 min US review
**Owner:** Offshore team lead
**US reviewer:** Marketing strategist
**Customer-facing SLA:** 5 business days from intake complete + payment received

## Required intake fields

From `ops/01-intake-form.md` — Section A (1-12), Section B (13-17 if branded template requested), and **Section C-1 (questions 18-26)**. All C-1 fields must be answered before production starts. Hard blocks:

- Q18 list CSV uploaded (or shared via Drive)
- Q22 goal of email
- Q23 key offer/message
- Q24 CTA URL (if applicable)

If any hard-block field is missing at Day 1 review: pause clock, ping US team to chase customer same day.

## Step-by-step execution

### Day 1 — Intake review (offshore, 30 min)
- [ ] Read full intake + kickoff call notes from US team
- [ ] Download customer list CSV; validate columns (email required; first_name/last_name optional)
- [ ] Run list through deduplication + syntax validation (remove obvious bad emails)
- [ ] Note list source (Q20) — if "purchased" flag for deliverability risk and notify US reviewer
- [ ] Confirm customer is on the sender authentication (SPF/DKIM aligned with our SendGrid sub-account, OR sending from `btwebgroup.com` on their behalf — confirm with US team)
- [ ] Set tracker status: `In production`

### Day 2 — Kickoff alignment (15 min)
- [ ] Review kickoff call notes; confirm brand voice notes (Q25) and preferred send window (Q26)
- [ ] Confirm production plan with US reviewer in Slack before drafting

### Day 3 — First draft (offshore + AI, ~1 hour)
- [ ] Use ChatGPT-4 or Claude to generate 3 subject-line variants and 1 preheader matched to goal (Q22) and offer (Q23)
- [ ] Generate first-draft body copy in customer's brand voice (Q25); include merge tag `{{first_name}}` where appropriate
- [ ] Pull or generate hero image (Midjourney or curated stock); ensure usage rights are clean
- [ ] Build email in SendGrid (or customer's ESP if migrating in); use mobile-first template; insert physical address + unsubscribe link (CAN-SPAM)
- [ ] Save working files to customer Drive folder: `/[CustomerName]/5K-Email-Blast/`

### Day 4 — Refinement + customer approval (offshore, ~45 min)
- [ ] Apply brand voice pass; trim length; verify single primary CTA points to Q24 URL
- [ ] Run internal QA checklist below
- [ ] Send test to internal QA inbox + mail-tester.com — must score ≥ 9/10
- [ ] Send PDF/preview to customer for written approval (non-negotiable — never send blast without)
- [ ] Set tracker status: `Ready for US review` once customer has approved

### Day 5 — US review, send, report
- [ ] US reviewer runs 04 checklist (email-send section)
- [ ] Schedule send for customer's preferred window (Q26)
- [ ] After send completes: pull 24-hour open/click report from SendGrid; build 1-page PDF results report (sends, opens, open rate, clicks, click rate, unsubscribes, top-clicked link)
- [ ] Send delivery email with PDF report + link to working files

## Internal QA checklist (run BEFORE flagging for US review)

- [ ] Subject line < 50 chars, no truncation in Gmail/Apple Mail preview
- [ ] Preheader < 100 chars, visible in preview
- [ ] All `{{first_name}}` merge tags resolve in test send (test with a row that has empty first_name — fallback should be graceful, e.g., "Hi there —")
- [ ] Mobile + desktop render verified (Litmus or Email on Acid, or manual on iPhone + Gmail web)
- [ ] Single primary CTA; CTA URL works and has UTM params (`utm_source=eblast&utm_medium=email&utm_campaign=[customer]-5k-blast`)
- [ ] Unsubscribe link present and functional
- [ ] Physical address in footer (CAN-SPAM)
- [ ] Mail-tester.com score ≥ 9/10
- [ ] No "AI-written" or "ChatGPT" mentioned anywhere in customer-facing copy
- [ ] Customer wrote "approved" in email/Slack — screenshot in tracker notes

## AI tools used (INTERNAL — never mention to customer)

- ChatGPT-4 or Claude — subject line variants, preheader, body first draft, brand-voice pass
- Midjourney or DALL-E — hero imagery alternatives when stock doesn't fit
- mail-tester.com — deliverability score
- SendGrid native preview — render check

## Common gotchas

- **Customer slow to provide list/login** — chase by end of Day 2; if not received by Day 3 morning, escalate to US lead
- **Customer brand voice unclear** — escalate to US lead for a 5-min direction call rather than guessing
- **Purchased list** (Q20) — flag deliverability risk to US reviewer immediately; may need to warn customer expected open rate will be low and recommend SEO/list-building upsell post-delivery
- **Customer never approves draft** — at Day 4 EOD if no approval, escalate to US lead; do NOT send without written approval
- **5-day SLA at risk** — escalate immediately, don't wait for it to actually miss

## Deliverable spec (per spec §4)

What the customer receives at delivery (matches spec §4 line 158):

1. 1 AI-drafted, team-edited email (HTML + plain-text fallback)
2. Subject line A/B test executed (winner sent to remaining list, or split 50/50 — document method in report)
3. Send to prospect's existing list, up to 5,000 contacts
4. Full delivery report (PDF): sends, opens, open rate, clicks, click rate, unsubs, top link
5. Working files in customer's Drive folder (HTML source, list as sent, report)

Cross-checked against spec §4: matches.