| name | outreach-sequence-builder |
| description | Build cold outreach sequences that get replies. Use when the user wants to write cold outreach emails, prospecting emails, cold email campaigns, sales development emails, SDR emails, or multi-channel outbound sequences. Also use when the user mentions 'cold outreach,' 'prospecting email,' 'outbound email,' 'email to leads,' 'reach out to prospects,' 'sales email,' 'follow-up email sequence,' 'nobody is replying to my emails,' 'how do I write a cold email,' 'cold email,' 'LinkedIn outreach,' 'outbound sequence,' or 'book more meetings.' Covers subject lines, opening lines, body copy, CTAs, personalization, multi-touch follow-up sequences, and multi-channel (email + LinkedIn + video). For warm/lifecycle email sequences, see email-sequence. For sales collateral beyond emails, see sales-enablement. |
| metadata | {"author":"Alfonso Sainz de Baranda (Growth4U)","version":"1.0","system":"SanchoCMO","phase":"Decide (one-to-one)","pillar":"outreach-sequence-builder","layer":"Decide","depends_on":"channel-prioritization, contact-enrichment, signal-definition","chains_to":"email-outreach-executor, linkedin-outreach-executor","context_required":["brand/{slug}/go-to-market/channel-plan.md","brand/{slug}/go-to-market/positioning/*/*.current.md","brand/{slug}/go-to-market/ecps/ecps.current.md","brand/{slug}/brand-voice/brand-voice.current.md","brand/{slug}/outreach-playbook/sequence-templates.md","brand/{slug}/outreach-playbook/discovery-guide.md"],"context_writes":["campaigns/","brand/{slug}/operational/learnings.md","brand/{slug}/operational/assets.md"]} |
Outreach Sequence Builder — Cold to Warm in 5-7 Touches
"Relevance beats volume. One perfectly-timed, signal-based email outperforms 100 spray-and-pray messages."
This skill creates personalized OUTBOUND cold sequences per ECP. It is NOT email-sequences (which handles INBOUND post-opt-in nurture). The distinction matters:
| Dimension | email-sequences (INBOUND) | THIS skill (OUTBOUND) |
|---|
| Audience | Opted-in subscribers | Cold prospects |
| Channels | Email only | Email + LinkedIn + video |
| Tone | Nurture, educational | Personalized, relevant, peer-to-peer |
| Trigger | Opt-in event (lead magnet) | Buy signal or ICP match |
| Goal | Convert subscriber → customer | Book meeting / demo |
| Personalization | Light (name, segment) | Deep (company, role, signal, pain) |
If the user asks for warm nurture sequences, redirect to email-sequences.
Read ./brand/ per _system/intelligence/brand-memory.md (if using SanchoCMO framework)
Follow _system/output/output-format.md (if using SanchoCMO framework)
Prerequisites
Required (will not run without these):
./brand/{slug}/go-to-market/channel-plan.md — Confirms outreach is a selected channel (from channel-prioritization)./brand/{slug}/go-to-market/ecps/ecps.current.md — Target personas with pain points (from niche-discovery-100x)./brand/{slug}/go-to-market/positioning/*/*.current.md — Value proposition angles per ECP (from positioning-messaging)
Recommended (better output with these):
./brand/{slug}/brand-voice/brand-voice.current.md — Tone for email copy (from brand-voice)- brand/{slug}/outreach-playbook/sequence-templates.md
- brand/{slug}/outreach-playbook/discovery-guide.md
brand/{slug}/operational/contacts-enriched.json — Enriched contacts for personalization (from contact-enrichment)brand/{slug}/operational/signals-to-track.json — Buy signals for triggers (from signal-definition)
Step 0: Tool Detection
Check environment for outreach tool availability:
Check for API keys / tool access:
INSTANTLY_API_KEY → Instantly (email sequencing)
LEMLIST_API_KEY → Lemlist (email + LinkedIn)
EXPANDI_API_KEY → Expandi (LinkedIn automation)
DRIPIFY_API_KEY → Dripify (LinkedIn automation)
FULL mode (2+ tools detected): Generate tool-ready sequences with platform-specific merge fields. Include import instructions.
STANDARD mode (1 tool or email client detected): Generate sequences optimized for that tool. LinkedIn touches are manual.
LIGHT mode (no tools): Generate copy-paste sequences. All touches are manual. Include timing calendar for manual tracking.
Announce detected mode to user. If LIGHT mode, note: "No outreach tools detected. Generating manual sequences. Consider Instantly (from EUR 30/mo) for email automation."
Workflow: 7 Steps
Step 1: Load ECP Context (~3 min)
For each ECP, extract from Context Lake:
From ecps.md:
- ECP name and description
- Pain points (ranked by intensity)
- Decision maker title
- Where they spend time online
- Buying triggers (what makes them look for solutions)
- Common objections
From positioning.md:
- Value proposition angle for THIS ECP
- Key differentiators vs alternatives
- Proof points (metrics, case studies)
From brand-voice/brand-voice.current.md (if exists):
- brand/{slug}/outreach-playbook/sequence-templates.md
- brand/{slug}/outreach-playbook/discovery-guide.md
- Tone guidelines
- Platform-specific adaptations
- Words to use / words to avoid
If the user specifies a single ECP, generate for that one. If no ECP specified, ask:
Which ECP should I build sequences for?
[1] [ECP 1 name] — [brief description]
[2] [ECP 2 name] — [brief description]
[3] All ECPs (generates one sequence per ECP)
Step 2: Load Signal Context (~2 min)
Read brand/{slug}/operational/signals-to-track.json if it exists. Map signals to sequence triggers:
Signal → Sequence Variant → Angle
──────────────────── ─ ──────────────────────── ─ ─────────────────────────
Fundraise → signal-fundraise.md → "Scale post-funding"
Hiring VP Marketing → signal-hiring.md → "Quick wins for new hire"
Tech stack change → signal-tech.md → "Maximize new tool ROI"
Competitor mention → signal-competitor.md → "Better alternative"
Content published → signal-content.md → "Continue the conversation"
If no signals-to-track.json exists, use the base sequence without signal variants. Note: "Signal variants improve reply rates 2-3x. Consider running signal-definition to define relevant signals."
Step 3: Select Sequence Structure (~2 min)
Present structure options from references/sequence-templates.md:
Sequence structure options:
[1] Multi-Channel Blitz — 7 touches, 14 days (email + LinkedIn + video)
Best for: high-value prospects, ABM-style
RECOMMENDED for most B2B outreach
[2] Value-First — 5 touches, 12 days (email + LinkedIn)
Best for: thought leadership brands, consultancies
[3] 3-3-3 Classic — 9 touches, 21 days (3 email + 3 LinkedIn + 3 follow-up)
Best for: enterprise, long sales cycles
[4] AIDA 4-Touch — 4 touches, 14 days (email only)
Best for: clear product, short sales cycle, SMB
Default to [1] Multi-Channel Blitz unless client context suggests otherwise.
Step 4: Define Personalization Variables (~3 min)
Using references/personalization-variables.md, determine personalization depth based on available data:
Level 1 (Basic): {first_name}, {company}, {title} — always available
Level 2 (Context): {industry}, {company_size}, {funding_stage}, {tech_stack} — from contact-enrichment
Level 3 (Signal): {recent_signal}, {trigger_event}, {pain_point_inferred} — from signal-monitor
Level 4 (Deep): {mutual_connection}, {content_they_published}, {custom_insight} — manual research
Present detected level: "Personalization Level 2 available (contacts-enriched.json found). For Level 3, run signal-monitor first."
Define fallback values for each variable used.
Step 5: Generate Sequence Per ECP (~10 min)
For each selected ECP, generate the full sequence following the chosen template structure.
Each touchpoint file includes:
# Touch [N]: [Channel] — [Name]
**Day**: [X]
**Channel**: Email / LinkedIn / Video
**Goal**: [What this touch achieves]
## Subject / LinkedIn Note
[Subject line with merge fields]
## Body
[Full copy with merge fields and fallbacks]
## Internal Notes
- Personalization level: [1-4]
- Merge fields used: [list]
- Fallback values: [list]
- Signal variant: [if applicable]
Copy rules:
- Email: Max 100 words per email (cold outreach is short)
- LinkedIn connection note: Max 300 characters
- LinkedIn DM: Max 150 words
- Always end with 1 clear CTA (never 2)
- Use the prospect's language (from ECP pain points), not your jargon
- First email: NEVER lead with "I" or "We" — lead with THEM
Step 6: GDPR Compliance Check (~2 min)
Present compliance checklist for the generated sequences:
GDPR Compliance Check:
[x] Legitimate interest basis documented
→ Reaching out about [specific business need relevant to their role]
[x] Opt-out mechanism in every email
→ "Not interested? Reply 'stop' and I'll remove you immediately."
[x] Professional context only
→ Using business email + LinkedIn (no personal data)
[x] Data source documented
→ Contact data from [Apollo/LinkedIn/company website]
[x] No purchased personal email addresses
→ All contacts sourced from professional databases
[ ] Data retention policy
→ Delete prospect data after 90 days of no engagement
→ ACTION: Set reminder in CRM/tool
If any item fails, flag it and suggest remediation.
Step 7: Present for Approval (~1 min)
Outreach sequence for [ECP name] ready:
[1] Accept (writes to ./campaigns/outreach-{ecp}/)
[2] Modify copy tone (more aggressive / more subtle)
[3] Change sequence structure
[4] Add/remove signal variants
[5] Generate for another ECP
Wait for user input before writing files.
Output
File Structure
./campaigns/outreach-{ecp-slug}/
brief.md ← Campaign overview + metrics targets
sequences/
01-cold-intro-email.md ← Day 0: Email
02-linkedin-connect.md ← Day 1: LinkedIn
03-value-email.md ← Day 3: Email
04-linkedin-engage.md ← Day 5: LinkedIn
05-proof-email.md ← Day 7: Email
06-linkedin-ask.md ← Day 10: LinkedIn DM
07-breakup-email.md ← Day 14: Email
signal-variants/ ← Optional (if signals defined)
signal-fundraise.md
signal-hiring.md
signal-tech.md
personalization-map.md ← Merge field definitions + fallbacks
brief.md Template
# Outreach Campaign — [ECP Name]
Generated: [date]
ECP: [name] — [description]
Sequence: [template name] — [N] touches over [N] days
Channels: [email, LinkedIn, video]
Personalization: Level [1-4]
Tool mode: [FULL/STANDARD/LIGHT]
## Target Metrics
- Open rate target: 40-60% (cold email benchmark)
- Reply rate target: 5-15%
- Meeting booked rate: 2-5%
- Sequence completion: < 3% unsubscribe
## Signal Variants
[list of signal variants included]
## GDPR Status
[compliance checklist summary]
Append summary to ./brand/{slug}/operational/assets.md:
[date] outreach-{ecp} — [N]-touch sequence ([channels]). Signal variants: [N].
Context Lake Integration
| Action | File | Description |
|---|
| READ | ./brand/{slug}/go-to-market/channel-plan.md | Confirms outreach is selected channel |
| READ | ./brand/{slug}/go-to-market/positioning/*/*.current.md | Value prop angle per ECP |
| READ | ./brand/{slug}/go-to-market/ecps/ecps.current.md | Target persona details |
| READ | ./brand/{slug}/brand-voice/brand-voice.current.md | Tone for email copy |
- brand/{slug}/outreach-playbook/sequence-templates.md
- brand/{slug}/outreach-playbook/discovery-guide.md
| READ |
brand/{slug}/operational/contacts-enriched.json | Enriched contacts (optional) |
| READ | brand/{slug}/operational/signals-to-track.json | Buy signals for triggers |
| WRITE | ./campaigns/outreach-{ecp}/ | Sequence files (multiple) |
| APPEND | ./brand/{slug}/operational/assets.md | Outreach sequence summary |
Frequency
- Per ECP: Create once, iterate monthly based on reply rates
- Signal variants: Add new variant when new signal type defined
- Refresh: When positioning changes or new case studies available
- A/B testing: After 50+ sends, create variant B for subject lines and opening hooks
Feedback Collection
After generating sequences, ask:
"Hay algun angulo de venta que funcione especialmente bien con este tipo de prospect? Alguna objecion comun que deberia anticipar?"
Log feedback to ./brand/{slug}/operational/learnings.md:
[date] outreach-sequence-builder: [ECP] — [feedback summary]
Plantillas de Partnerships (SAN-80) — biblioteca + asignación + gates
Para campañas de creators (Outreach·Partnerships) las secuencias y briefs viven como
plantillas-asset en brand/{slug}/outreach/templates/*.md (frontmatter + pasos
## Paso N · Título (espera Xd) con **Asunto:**). El catálogo de variables
NO se hardcodea en el prompt: se lee de GET /api/partnerships/templates?slug={slug}
(campo variables). Cada entrada declara token, sourcePath y example.
Acciones disponibles (paridad UI = chat = MCP — usa SIEMPRE estos endpoints, no toques los .md a mano):
| Acción | Endpoint | MCP tool |
|---|
| Listar biblioteca | GET /api/partnerships/templates?slug={slug} | — |
| Crear/editar plantilla | POST /api/partnerships/templates · PUT /api/partnerships/templates/{id} | — |
| Asignar (instanciar copia) a una búsqueda | POST /api/partnerships/templates/{id}/assign {searchId|campaignId} | yalc_assign_template |
| Contactar creators (instancia secuencia → GateItem) | POST /api/partnerships/contact {leads:[{id,campaignId}]} | — (gate vía tools) |
| Listar gates pendientes | GET /api/yalc/gates?slug={slug} | yalc_list_gates |
| Aprobar/rechazar gate de envío | POST /api/yalc/gates {runId, action} | yalc_approve_gate |
Reglas:
- La biblioteca guarda originales; cada búsqueda congela copias (instancias). Editar el original NO cambia búsquedas ya asignadas.
- Usa únicamente tokens presentes en
variables. Nunca inventes placeholders (anchor_topic, anchor_specific, etc.) ni derives su valor mediante LLM/scraping. Si falta el campo, reescribe el copy sin esa variable.
- Antes de guardar, confirma que cada
{{variable}} aparece en el catálogo. La API rechazará cualquier variable desconocida.
- Usa solo la forma exacta
{{variable}}; fallbacks y sintaxis alternativa no son compatibles con el renderer de envío.
- El plan de discovery puede traer la fila
templates: [nombres] — createDiscoverySearch las asigna solo si existen en la biblioteca; ofrece crear las que falten.
- El envío SIEMPRE pasa por el gate humano (dry-run por defecto: jamás un email real sin
dryRun:false explícito y proveedor configurado). No apruebes gates sin instrucción del usuario.
