Dual Wing Rental
Marketing Engine

SEO M1.1 — SEO Acquisition Subsystem

M1.2 — Social Content Engine

Content Multiplication System

INPUT  : 1 content idea (row in Google Sheet)
OUTPUT :
  1 long-form blog post (800–1200 words, SEO optimised)
  3 short-form video scripts (TikTok / Reels / Shorts)
  2 carousel posts (Instagram / Facebook)
  1 email (nurture list)
  1 retargeting ad angle
  3 caption variants (short / medium / long with hashtags)

WORKFLOW (n8n + Claude API):
  TRIGGER : New row added to "Content Ideas" Google Sheet
  STEP 1  : n8n reads row → [Topic / Wing / Keyword / Priority]
  STEP 2  : n8n sends to Claude API with full prompt (see M11.1)
  STEP 3  : Claude returns structured content JSON
  STEP 4  : n8n saves outputs to Google Drive
  STEP 5  : Make sends notification with Drive link
  STEP 6  : Human review gate — MANDATORY before any publish
  STEP 7  : Approved → Buffer scheduled
  STEP 8  : Approved video → Repurpose.io queued

Hook Bank — Landlord

1. "Your [city] property could earn £/$/X more per month"
2. "3 mistakes landlords make that cost them tenants"
3. "How we filled a property in [city] in 4 days"
4. "What Airbnb doesn't tell you about your rental income"
5. "Empty property = losing £/$/X every single day"
6. "Why your property isn't getting inquiries (honest answer)"
7. "Guaranteed rent vs Airbnb: which actually pays more?"

Hook Bank — Renter

1. "Affordable apartments in [city] that are actually verified"
2. "How to get approved for a rental in [city] fast"
3. "This studio in [city] had 11 inquiries in 24 hours"
4. "Move-in ready apartments in [city] — no hidden fees"
5. "The honest guide to renting in [city] in 2026"

Posting Schedule

M1.3 — Listing Platform Layer

UK PRIORITY ORDER:
  Tier 1 (launch):   Airbnb, Booking.com, Facebook Marketplace
  Tier 2 (Month 3+): SpareRoom, Gumtree, Zoopla/Rightmove*
  // *Rightmove/Zoopla API: 4–8 weeks minimum, gatekept. Plan for manual CSV first.

US PRIORITY ORDER:
  Tier 1 (launch):   Airbnb, Booking.com, Facebook Marketplace, Craigslist
  Tier 2 (Month 3+): Zillow Rental Manager, Apartments.com, Zumper

⚠ KNOWN GAP: Two-way sync is not fully solved.
  A booking on Airbnb will NOT update GHL automatically without
  Hospitable ($40/mo) or Guesty (enterprise).
  First 90 days: external platform calendars = source of truth for availability.
  GHL = source of truth for leads and contacts. Document this clearly.

M1.4 — Paid Traffic (Optional Accelerator)

FACEBOOK / INSTAGRAM STRUCTURE (1 campaign per wing):
  Ad Set 1 — COLD (40% budget)
    Audience: Lookalike or interest-based
    CTA: Learn More → landing page

  Ad Set 2 — WARM (30% budget)
    Audience: Page visitors last 30 days
    Creative: proof (case study, earnings, reviews)
    CTA: Get Projection / Book Viewing

  Ad Set 3 — HOT (30% budget)
    Audience: Form starters who didn't submit / 75%+ video viewers
    Creative: urgency / scarcity
    CTA: Direct form or WhatsApp

// 2026: Enable Meta CAPI + TikTok Events API via GHL native integration.
// Without server-side events, you're working with 30–50% of signal.

LANDLORD PAGE FLOW:
  URL:       /rent-out-property-[city]
  Form  →    GHL pipeline "Landlord - New Lead"
  Thank you: /landlord-next-steps (what happens next + calendar embed)

RENTER PAGE FLOW:
  URL:       /rentals-[city]
  Form  →    GHL pipeline "Renter - New Inquiry"
  Thank you: /renter-next-steps (available listings + WhatsApp CTA)

MICRO-FORM (reduce abandonment 30–50%):
  Above fold: email + city only (2 fields)
  Step 2: full form reveals after submission
  Use for cold traffic. Full form for warm/retargeting.

SUBMISSION TRIGGER (GHL Workflow):
  ACTION 1: Create contact record
            → Check for duplicate by phone number FIRST
            → If duplicate: merge or flag — do NOT create second record
  ACTION 2: Apply tag [Wing: Landlord] or [Wing: Renter]
  ACTION 3: Apply tag [Source: {{utm_source}}]
  ACTION 4: Create opportunity in correct pipeline
  ACTION 5: Check suppression list — if suppressed, stop
  ACTION 6: Fire automation sequence (Module 5)

⚠ GHL OUTAGE SAFETY NET (5 min setup, zero cost):
  Configure backup email notification to personal inbox on form submit.
  Fires independently of GHL automation. Catches leads during GHL downtime.

Score Buckets

Landlord Scoring Matrix

Lead Score Decay (n8n — runs nightly)

// A static score is lying to you.
// A lead scored 70 on Day 0 is not a 70 lead on Day 7 with no engagement.

TRIGGER: n8n scheduled — midnight, every night
STEP 1:  Pull contacts with lead_score > 0 AND tag [Sequence Active]
STEP 2:  For each contact:
           IF days_since_last_engagement > 7:
             lead_score = lead_score - 15
             Apply tag [Engagement: Cooling]
           IF days_since_last_engagement > 14:
             lead_score = lead_score - 20 additional
             Apply tag [Engagement: Cold]
             Move to COLD sequence if currently WARM
STEP 3:  Write updated score → GHL custom field
STEP 4:  Log adjustment in GHL contact notes

ENGAGEMENT EVENTS THAT RESET DECAY CLOCK:
  Email open     → +5   (mild interest)
  Email click    → +10  (strong interest)
  SMS reply      → +15  (active engagement)
  Page visit     → +10  (return intent)
  Form partial   → +15  (high intent, incomplete)

Landlord Pipeline Stages

Renter Pipeline Stages

GHL Custom Fields

LANDLORD FIELDS:
  property_type, property_city, property_bedrooms, currently_occupied
  current_monthly_income, expected_income
  lead_score              // written by n8n
  lead_temperature        // HOT/WARM/COLD — written by n8n
  utm_source, utm_medium, utm_campaign  // hidden form fields
  sequence_day, last_contacted_date, last_engagement_date
  days_since_engagement   // calculated nightly by n8n
  contract_signed_date, property_listed_date
  referred_by             // contact ID
  human_review_required   // Y/N — set for HOT leads

RENTER FIELDS:
  budget_min, budget_max, move_in_date, property_type_wanted
  city_preference, num_occupants
  lead_score, lead_temperature
  listings_sent, listings_matched_count  // written by n8n inventory check
  viewing_date, booking_date
  saved_search_criteria   // for Notify Me / waitlist feature
  referred_by

Landlord Sequence — Days 0–7

Renter Sequence — Days 0–7

Behavioral Triggers (Interrupt Any Sequence)

Contact replies to any message:
  → Pause sequence immediately
  → Task: "Reply received — manual follow up now"
  → Tag [Replied] · Move stage → Contacted

Contact clicks booking link but doesn't complete:
  → SMS within 30 min: "Looks like you started booking — anything I can help with?"

Contact opens email 3+ times without responding:
  → Tag [High Interest - No Action]
  → Task: "High email engagement — call now"
  → Trigger AI personalized follow-up (M11.2)

Contact sends keyword "EMERGENCY", "ASAP", "urgent", "call me now":
  → Tag [Emergency: Urgent Contact]
  → Immediate Slack alert to owner
  → SMS from owner number within 15 min

New listing added matching renter's saved search criteria:
  → SMS: "We found it — a [property type] in [city] just became available"
  → Email with property details

LANDLORD CONSULTATION BOOKING (GHL calendar triggered):
  Email (immediate): confirmation + what to expect + calendar links
  SMS  (immediate): confirmed + reschedule option
  SMS  (24h before): reminder
  SMS  (1h before):  reminder

POST-CONSULTATION (stage moved to Evaluated):
  Email (within 2h of call): personalised revenue projection
  ┌──────────────────────────────────────────────┐
  │ Estimated monthly income : £/$/[X]–[X]       │
  │ Management fee           : [X]%              │
  │ Net to you monthly       : £/$/[X]–[X]       │
  └──────────────────────────────────────────────┘
  CTA: [VIEW AGREEMENT]

AGREEMENT FOLLOW-UP (48h no signature):
  Email + SMS: "Any questions about the agreement?"

AGREEMENT SIGNED (DocuSign/PandaDoc webhook → GHL):
  → Move stage: Closed - Won
  → Fire onboarding workflow (M7)
  → Create property onboarding task list

RENTER — INTENT TO BOOK CONFIRMED:
  Email: Step 1 application form + Step 2 holding deposit payment link
  SMS: "I've reserved [property] for 24 hours. Complete your application: [link]"

PAYMENT RECEIVED (Stripe webhook → GHL):
  → Move stage: Closed - Won
  → Fire fulfillment workflow (M7)

LANDLORD ONBOARDING (triggered: Stage = Closed - Won):
  STEP 1: GHL task list created
    ☐ Request property photos + access
    ☐ Collect docs: EPC (UK) / Inspection (US), gas cert, insurance
    ☐ Management agreement signed
    ☐ Property set up in system
    ☐ Listing assets created
  STEP 2: Welcome email → secure upload link + photo guidelines
  STEP 3: Make workflow polls for document uploads → team notified on completion
  STEP 4: Listing live confirmation (Email + SMS)
  STEP 5: 7-day check-in: views / inquiries / bookings report

RENTER STAY MANAGEMENT (booking confirmed + payment):
  Day 0:    Full booking confirmation email (check-in date, key instructions)
  Day -3:   SMS: "Your stay starts in 3 days!"
  Day 0 AM: SMS + Email: check-in instructions (key code, WiFi, parking)
  Day 0 +2h: SMS: "Hope you're settled in — anything needed?"
  Day -1:   Checkout reminder (time + key return)
  Day +1:   Review request (Google / Airbnb / Booking.com links)

LANDLORD RETENTION:
  Monthly automated report (Make + GHL, 1st of month):
    Bookings · Occupancy rate % · Revenue · vs. last month · Upcoming bookings

  Upsell — IF occupancy < 70% for 2 consecutive months:
    → Optimisation consultation email offer
  Upsell — IF landlord has 1 property AND 6+ months active:
    → Multi-property upsell sequence

RENTER RETENTION:
  Day 1:  Review request
  Day 7:  "We'd love to have you back" + 10% returning guest discount
  Day 30: "Any upcoming plans?" + availability CTA

  Loyalty tags:
    After 2 stays → [Returning Guest]
    After 3 stays → [Loyal Guest] + priority support

LANDLORD REFERRAL (triggered: Closed - Won + 30 days):
  Email: "Know another landlord in [city]?" + reward details
  SMS (1 week later, no referral): reminder

  GHL tracking:
    custom field: referred_by (contact ID)
    tag [Referred Lead] on new contact
    tag [Referring Contact] on referrer
    On referral close → issue reward + thank you email

RENTER REFERRAL (triggered: positive review 3+ stars):
  SMS +24h: "If friends are visiting [city] — they get 10% off, you get 10% off"
  Email: referral program details + shareable tracking link

FACEBOOK / INSTAGRAM AUDIENCES:
  Audience 1: All website visitors (180 days)
  Audience 2: Landlord landing page visitors (180 days)
  Audience 3: Renter landing page visitors (180 days)
  Audience 4: Form starters — did not submit (90 days)
  Audience 5: Thank you page visitors = converters [EXCLUDE FROM ALL AD SETS]
  Audience 6: Video viewers 75%+ (180 days)

  Lookalike audiences: Build ONLY after 100+ conversions

EXCLUSION SYNC (Make):
  TRIGGER : Contact tagged [Converted]
  ACTION  : Make adds email to Facebook Custom Audience exclusion list
            Prevents serving ads to existing customers. Saves budget.

// 2026: Enable Meta CAPI + TikTok Events API via GHL.
// Without server-side events: 30–50% of signal. LAL quality degrades.

Automated Content Generation Workflow

TRIGGER: New row in "Content Ideas" Google Sheet
  Columns: [Topic] [Wing] [Target Keyword] [Priority] [City]

STEP 1: n8n reads row
STEP 2: n8n sends to Claude API (Sonnet)

PROMPT TEMPLATE:
────────────────────────────────────────────────────────
"You are a content writer for a rental property management company.
Topic: [topic]  Target audience: [Wing]  Keyword: [keyword]
City: [city]  Market: [UK/US]
Tone: Conversational, trustworthy, direct. No fluff. Specific.

Generate:
1. BLOG POST — 800–1000 words, H1/H2, keyword 3–4x naturally, CTA at end
2. VIDEO SCRIPT 1 — 60 seconds, hook/value/proof/CTA
3. VIDEO SCRIPT 2 — 30 seconds, urgency/curiosity
4. VIDEO SCRIPT 3 — 90 seconds, educational/case study
5. EMAIL — 150–200 words, subject + single CTA button label
6. CAROUSEL — 6 slides: slide 1 hook, 2–5 value, 6 CTA
7. CAPTIONS — short / medium / long with hashtags (max 10)

Return ONLY valid JSON: {blog, video1, video2, video3, email, carousel, captions}"
────────────────────────────────────────────────────────

STEP 3: Claude returns JSON
STEP 4: n8n parses JSON
        // try/catch — Claude occasionally returns malformed JSON
        // on error: log, retry once, alert if fails twice
STEP 5: n8n saves to Google Drive
STEP 6: Sheet row status → "Generated"
STEP 7: Make → Slack/email: "✅ Content ready: [Topic] — [Drive link]"
STEP 8: Human review — MANDATORY gate
STEP 9: Approved → Buffer scheduled · Approved video → Repurpose.io

AI Personalized Follow-Up (n8n + Claude API)

TRIGGER: Contact tagged [High Interest - No Action]

STEP 1: n8n pulls contact data from GHL API
  Fields: name, wing, city, property_type, budget, source,
          days_since_entry, sequence_day, lead_score

STEP 2: n8n sends to Claude API
  "Write a personalized follow-up SMS (max 160 chars)
   and email (subject + 100 words max) for this lead.
   Rules:
   - Do NOT use: 'just checking in', 'hope this finds you well',
     'circling back', 'touching base'
   - DO include: one specific reference to their city or property type
   - Do not fabricate urgency. Only reference real scarcity if confirmed.
   Return ONLY valid JSON: {sms, email_subject, email_body, email_cta}"

STEP 3–7: Send SMS via GHL API · Send email via GHL API · Log in timeline

Looker Studio Dashboard Structure

Weekly Intelligence Report (n8n + Claude API)

TRIGGER: Every Monday 8:00am

STEP 1: n8n pulls last 7 days from Google Sheets
        (leads/channel · conversion rates · email open/click · HOT leads actioned)

STEP 2: n8n sends to Claude API
  "You are a growth analyst for a dual-wing rental marketing business.
   Here is last week's performance data: [data]
   In 200 words max:
   1. What worked — top 2 things, with numbers
   2. What needs fixing — top 2 things, specific cause if identifiable
   3. This week's priority action — 1 specific, executable thing
   Rules: Be direct. If data is insufficient, say so.
   Do not invent trends from sparse data.
   Flag any metric that looks like a tracking error."

STEP 3: Claude returns analysis
STEP 4: n8n sends → email + Slack: "📊 Weekly Marketing Intel — [date]"

GOOGLE SHEETS STRUCTURE:

Table 1: Cost Per Lead
  [Channel] [Spend] [Leads] [CPL] [Notes]
  SEO (time) → log hours × rate. Not zero.

Table 2: Cost Per Acquisition
  [Wing] [Leads] [Closed] [Close %] [CPA]

Table 3: Revenue Per Property
  [Property] [Mgmt Fee %] [Avg Booking Rev] [Net/Month]

Table 4: Monthly Tool Costs (Actual vs Budget)
  GHL · Make · n8n VPS · Claude API · Canva · Buffer · other
  TOTAL: ~$168–430/mo depending on tier

Table 5: Operator Time Ledger
  [Activity] [Hours/Week] [Hourly Rate] [Weekly Cost]

DECISION RULES:
  IF CPL > threshold AND close rate < X%:
    → Pause or kill that channel
  IF tool cost > attributable revenue generated:
    → Evaluate replacement or removal
  IF operator time > 30% of revenue:
    → Identify highest-time tasks for automation priority
  IF HOT lead conversion rate < 50%:
    → Human response protocol is broken — audit immediately

FLYWHEEL LOOP:
STEP 1: Landlord closes → property listed on platforms
STEP 2: New listing → n8n → Claude API
          → generate 5 hooks specific to this property
          → generate 3 listing-specific social posts
          → schedule via Buffer automatically
STEP 3: Content distributed → renter inquiries increase
STEP 4: Renter inquiries captured in GHL → tagged by city
STEP 5: n8n generates weekly "demand proof" report by city
STEP 6: Demand proof email → all WARM landlord leads in that city
          "We had [X] renter inquiries in [city] this week.
          Is your property ready to capture this demand?"
STEP 7: Social proof raises landlord conversion rate
STEP 8: More landlords → more listings → back to STEP 2

FLYWHEEL BREAKS IF:
  ⛔ Landlord supply drops (acquisition stalls)
  ⛔ Renter inventory is empty (no social proof for landlords)
  ⛔ Demand proof email isn't sending (Make workflow offline)
  ⛔ Demand threshold too high (no emails fire)

DEMAND PROOF WORKFLOW (Make — every Monday 9:00am):
  STEP 1: Pull last 7 days renter inquiries from GHL by city
  STEP 2: IF inquiries in city > [threshold — start at 3]:
            Pull all WARM landlord leads in same city
            Claude API → generate demand proof email
            Send via GHL to segment
  STEP 3: Log open rate + click rate to Google Sheets
  STEP 4: IF threshold not met → log "Insufficient demand data — [city]"
            Do NOT send. Empty proof emails kill credibility.

n8n (self-hosted VPS) — UptimeRobot (free):
  Ping n8n webhook endpoint every 5 minutes
  Alert: SMS to owner + email if down
  Auto-restart via systemd service on VPS
  Setup time: 20 minutes. Non-negotiable.

GHL:
  Notable outages in 2023, 2024, 2025.
  Backup: on every form submit, direct email to personal inbox fires
  independently of GHL automation. Catches leads during GHL downtime.
  Cost: zero. Setup: 5 minutes.

Make / Zapier:
  Set: notify immediately on scenario failure.
  Review error logs weekly — silent failures are common.

Claude API:
  n8n: try/catch on all Claude API calls
  On error: log to Sheet, send Slack alert, skip (don't halt)
  Content machine failing silently for a week is a real risk.

Weekly System Health Check (Every Friday — 30 min)

FORMS:
  ☐ Submit test form (both wings)
  ☐ Verify GHL contact created + lead score populated + tags applied
  ☐ Verify Day 0 SMS + email fired within 5 min

SEQUENCES:
  ☐ Last 10 new contacts — did sequences fire correctly?
  ☐ Behavioral triggers working? (reply → pause sequence)

CONTENT MACHINE:
  ☐ Add test row to Content Ideas sheet
  ☐ Verify n8n triggers and Drive files created

ANALYTICS:
  ☐ GA4 receiving data (realtime check)
  ☐ Looker Studio connections not broken
  ☐ Unit economics sheet updated

FLYWHEEL:
  ☐ Monday demand proof email sent (check GHL sent log)
  ☐ Weekly Intel Report arrived Monday 8am

MONITORING:
  ☐ UptimeRobot — any downtime events?
  ☐ Make scenario error log — any failures?
  ☐ Claude API usage — tracking against budget?

Automation Platforms

AI Content Generation

WhatsApp Channel Setup

PLATFORM: GHL WhatsApp Business API
  Requires: Meta Business verification (1–3 business days)

USE CASES IN THIS SYSTEM:
  Renter Day 0: Listing photos via WhatsApp (higher engagement than email)
  HOT Lead Alert: Owner notified via WhatsApp (faster than email)
  Viewing Reminders: Day before + 2hr before
  Post-Stay: Review request (higher open rate than SMS)

// UK 2026: WhatsApp open rates 90%+ vs SMS 82% vs email 20–35%
// Build in Week 1 for UK-focused. Do not treat as optional.
// US: SMS still dominant. WhatsApp for international communities in
// major metros (NYC, Miami, LA, Houston).