GolfOkay Knowledge Vault

hybrid_integration

5 min read

Hybrid Integration

GolfOkay’s core technical strategy: Accept that many Thai golf courses will never have APIs. Build for that reality.

The Reality

Course digital readiness (working estimates):

  • Minority: Modern systems with APIs (Golfmanager, Lightspeed, etc.)
  • Some: Basic websites, no real-time booking
  • Many: Manual only (phone, LINE, email)

Not a failure. This is the market. Waiting for full automation = waiting forever.

Three Integration Tiers

Tier 1: Aggregator APIs (Immediate)

Sources: Golfsavers, GolfThai

Coverage: 200-500 courses

Speed: Real-time availability, instant confirmation

Development time: 2-4 weeks

Use case: Tourist favorites, popular courses, championship venues

Pros:

  • Live from day 1
  • Instant booking confirmation
  • Standardized workflow

Cons:

  • Aggregator commission (3-5%)
  • Can’t negotiate custom rates
  • Still miss 50% of market

Tier 2: Direct Course APIs (Selective)

Sources: Courses using Golfmanager, Lightspeed, Jonas Club Software

Coverage: 40-60 courses (15-20% of market)

Speed: Real-time for API courses

Development time: 3-5 days per course system

Use case: High-volume courses where direct relationship justifies integration effort

Pros:

  • No aggregator fees
  • Better rates possible
  • Direct relationship with course

Cons:

  • Each course system different
  • Maintenance overhead
  • Limited to tech-capable courses

Tier 3: Manual Desk (Essential)

Sources: All non-API courses (50-55% of market)

Coverage: 150-170 courses

Speed: <48hr confirmation (SLA target)

Operations: Human staff calling/LINE messaging courses

Use case: Local gems, expat favorites, budget courses not on aggregators

Pros:

  • Access to courses competitors can’t reach
  • Flexibility for custom packages
  • Personal relationships with course staff

Cons:

  • Labor cost (~฿200-300 per booking)
  • Slower than API (24-48hr vs instant)
  • Scales with headcount, not just code

See manual_desk for detailed workflow.

Platform Architecture

Customer inquiry (LINE bot / web)
   ↓
Intent classification (AI)
   ↓
Course availability check:
   ├─> Tier 1: Check Golfsavers API
   ├─> Tier 1: Check GolfThai API
   ├─> Tier 2: Check direct course API (if integrated)
   └─> Tier 3: Route to manual desk
   ↓
Quote generation (<24hr per [[sla|SLA]])
   ↓
Customer accepts
   ↓
Payment via [[promptpay|PromptPay]] / card
   ↓
Booking confirmation
   ↓
[[e_vouchers|E-voucher]] generation

Unified customer experience: Whether booked via API or manual desk, customer gets same e-voucher format and service level.

Automation Targets

Month 3:

  • 50-60% of inquiries fully automated (FAQ, quote generation)
  • 30-40% require manual desk involvement
  • <24hr quote SLA maintained

Month 6:

  • 65-70% of bookings via Tier 1/2 (API-enabled)
  • 30-35% via manual desk
  • CSAT ≥4.6/5 across all tiers

Month 12:

  • 70-75% API-enabled (add direct integrations for high-volume courses)
  • 25-30% manual desk (steady state)
  • Manual desk efficiency: 15-20 bookings/day per operator

Why Hybrid Works

Market reality: Can’t API-integrate every course. They don’t have systems.

Customer expectation: Doesn’t matter how booking happens. They want their tee time confirmed.

Competitive advantage: Competitors either go full-API (miss 50% of courses) or full-manual (can’t scale).

GolfOkay does both: API where possible, manual where necessary.

Manual Desk as Asset, Not Liability

Traditional view: Manual processes = failure to automate

Reality: Manual desk enables access to inventory competitors can’t reach

Examples:

  • Hidden gem course in Hua Hin (no website, LINE bookings only)
  • Exclusive member course (referral only, phone bookings)
  • Local favorite with expat community (cheap, manual-only)

These courses matter. Expats know them. Reviews are excellent. But they’ll never have APIs.

Operational Efficiency

Manual desk cost: ฿200-300 per booking (15-20 min × ฿50/hr loaded cost)

Tier 1/2 cost: Negligible (API call + payment processing)

Break-even: If manual desk books 100-150 packages/month, operational cost is ฿20-45K/month

Revenue from those bookings: 100 packages × ฿25K margin = ฿2.5M revenue

Manual desk cost as % of revenue: 0.8-1.8% (acceptable overhead for 30% of market access)

Scaling Manual Desk

Month 0-3: Tan handles manual desk solo (20-30 bookings/month capacity)

Month 4-6: Hire ops manager (฿30K/month) – capacity doubles to 50-60 bookings/month

Month 7-9: Add 2nd operator if volume exceeds 80 bookings/month

Month 10-12: 3-person manual desk = 150-180 bookings/month capacity

Efficiency improvements:

  • Pre-negotiated time windows with courses (reduces coordination)
  • Templated LINE messages (faster communication)
  • AI-assisted intent classification (routes inquiries correctly)
  • Macros for common scenarios (quote generation)

White-Label Course Tools (Future)

Phase 2 strategy (Month 10+):

  • Offer free booking engine to manual-only courses
  • “We’ll build you a booking system, you give us direct access”
  • Converts Tier 3 courses to Tier 2 over time
  • Trade technology for inventory access

Value prop for courses:

  • Free modern booking UI
  • Mobile-friendly
  • Integrated payment processing
  • Real-time availability management

Value for GolfOkay:

  • Direct API access (no aggregator fees)
  • Exclusive rates possible
  • Reduces manual desk load

Technology Stack

API layer:

  • Node.js/TypeScript backend
  • REST API wrapper for Golfsavers, GolfThai
  • GraphQL for direct course integrations (flexible schema per course)

Manual desk tools:

  • CRM for tracking inquiries (status: pending/quoted/confirmed/completed)
  • LINE integration for customer communication
  • SMS alerts for operators when inquiry routed
  • Dashboard showing manual desk queue

AI automation:

  • Intent classification (golf inquiry vs general question)
  • Course recommendation engine (budget, location, skill level)
  • Quote generation (pull rates, calculate package price)
  • Follow-up automation (remind customer of pending quote)

Competitive Context

Golfasian: 100% manual (email/phone). Slow but comprehensive coverage.

Golfsavers: 100% API (aggregator model). Fast but limited to API-enabled courses.

GolfOkay: Hybrid. Fast for 70% of courses, comprehensive for 100% of courses.

Customer doesn’t care whether booking was automated or manual. They care that it works.

See Also

Graph View

Backlinks