1. The situation
A homestay network in Trang province, spread across multiple areas, was founded as a cooperative network in 2018 by a local NGO promoting community-based tourism. Room rates averaged THB 850-1,800/night including breakfast + activities (rubber tapping at 4 AM, kitchen garden harvesting, traditional snack making). The pain point shared during discovery with the network board and several homestay owners: 88% of revenue came from Airbnb, which deducted 14-17% commission + 2024 Airbnb policy updates reducing host control led to some owners being deactivated without clear reason (several were deactivated during high season → revenue lost).
More importantly, Airbnb’s algorithm mixed homestays nationwide — anyone searching “Trang homestay” saw the network’s listings alongside independent listings + hotels listing on Airbnb as stand-alone — meaning the Trang cluster’s distinctive proposition (community network + organic farming + halal-friendly + 95% revenue returning to community) lost its competitive advantage.
Beyond that, the board saw a major opportunity in Muslim tourism from Malaysia + Singapore — Trang has 18% Muslim population + distinctive southern halal cuisine (roti, khao mok, khao yam, cha) + only 4 hours’ drive from Penang. But only 8% of customers were Muslim because Airbnb has no halal-certified filter, no qibla direction info, no Malay-language content explaining which homestays are halal-friendly.
The goal set at kickoff: a shared booking platform for the homestay network that reduced Airbnb dependency from 88% to no more than 55% + grew Muslim tourist mix from 8% to 30%.
2. Why off-the-shelf solutions didn’t work
Each homestay had attempted its own website — some used Wix, some used a Facebook Page as the “website”, some had nothing. The problem: no shared availability — customers wanting to book a network homestay had to browse each one and check availability one at a time. If their first choice was full they had to look at the next one. UX was terrible.
We talked with Hostfully + Lodgify + iGMS, property management software for multi-listing rentals. License fees ran USD 50-150/month per property — for the network the cost was high + 5-year TCO of USD 24,000-72,000 — unsustainable overhead for rural Trang homestays averaging THB 28,000-45,000/month per property.
Beyond cost, these SaaS were designed for Western “vacation rental” workflows, not the “community cooperative” business logic the network needed — automatic revenue split (owner 75% + cooperative fund 20% + admin fee 5%), shared activity booking where a customer chose homestay A + joined rubber tapping at homestay C in a single transaction, multi-property dietary preference matching (Muslim customers automatically recommended halal-friendly homestays).
Worse, none of those SaaS supported PromptPay QR (required for Thai customers), and activating Stripe under a cooperative entity requires Stripe Atlas (incorporating a Delaware company) — too complex for a community organisation.
3. Our approach
After 3 rounds of discovery — board meeting + on-site visits to several homestays (eating, sleeping, waking up to tap rubber at 4 AM) + workshop with the full network — we planned 4 pillars.
1) Cooperative website under network brand. Built a single website featuring the entire homestay network under a clear network brand. Customers could search + browse + compare + book in one place. Used Astro for LCP <1.5s + Cloudflare Pages free + R2 for 320+ photos of the network.
2) Custom booking engine handling cooperative business logic. Built on Supabase: shared availability across all properties in one calendar, automatic revenue split (owner 75% + fund 20% + admin 5%), shared activity booking where a customer booked homestay A + activity at homestay C in one transaction, payment via PromptPay QR (Thai) + Stripe (foreign, through a cooperative account registered as a Thai social enterprise).
3) Halal-friendly content + Mandarin/Malay SEO. Built a TH/EN/MS content hub of 16 articles covering “Halal-friendly homestay in Trang”, “Trang Muslim food guide”, rubber tapping experience, dugong watching in Trang province, organic farm tour. Each property tagged “halal-certified” / “halal-friendly” / “general” honestly — implemented Schema HalalCertified + LodgingBusiness + qibla direction in room info.
4) Local SEO + GBP listings + dugong tourism. Optimised Google Business Profiles per property (each is an independent GBP linked back to the network), 30+ local citations, content cluster about dugong watching in Trang (the best dugong viewing in Thailand, Trang’s geographic moat).
4. Week-by-week
Week 1-2: Discovery + Cooperative agreement. 2-day workshop with the entire network of owners + the board → aligned revenue split formula, drafted cooperative website Terms of Use, designed a 16-table database schema covering property, room, activity, customer, booking, payment, revenue distribution.
Week 3-5: Backend + Booking Engine. Built booking API on Supabase, shared availability calendar across all properties, revenue split automation (atomic transaction that splits + updates accounting log + notifies property owner), payment integration PromptPay + Stripe + bank transfer.
Week 6-7: Frontend + Multilingual. Astro 28 pages including cooperative landing + property detail + booking flow, i18n routing TH/EN/MS, mobile-first design (78% of traffic was mobile), halal-friendly UX (qibla direction badge, prayer time widget, halal-certified property filter).
Week 8: Content + SEO + GBP. Wrote 16 articles TH/EN/MS (native writers in all 3 languages), implemented full Schema markup, set up GBP listings with professional photos + business descriptions, submitted citations to 32 directories including halal-specific (HalalTrip, CrescentRating, Halal Booking).
Week 9: QA + Launch. End-to-end booking flow test including payment + revenue split, WCAG AA accessibility audit, soft launch timed to coincide with Malaysia/Singapore Hari Raya holiday to test Muslim tourist response.
After launch the 6-month retainer began: 2 articles/month + weekly GBP posts for 8 listings + monthly performance review.
5. Obstacles + Pivots
Obstacle 1: Some owners rejected the revenue split formula. Week 1 we discovered some owners felt the 20% cooperative fund was too high (it had been 12% before). Fixed via a transparency workshop — showed how the cooperative fund covered shared marketing + property maintenance + annual Muslim accreditation costs. Reducing to 12% wouldn’t cover certification fees + agreed split was finalised at 75/18/7.
Obstacle 2: Halal certification process took longer than expected. The network wanted several properties to be halal-certified by the Central Islamic Council of Thailand (CICOT) by launch, but the inspection + approval process takes 4-6 months. Fixed by launching with “halal-friendly” status (food and lodging accommodate Muslim guests but lack formal certification) + clearly stated on the website + planned upgrade to “halal-certified” within 6 months post-launch.
Obstacle 3: PromptPay slip reconciliation manual workload. In month 1 the network’s admin had to verify PromptPay slips against bookings one by one — 3 hours/day. Fixed by integrating Kasikorn Bank’s Slip Verify API + auto-matching booking amount + reference — workload dropped 88%.
6. Post-launch + Ongoing
Within 5 months post-launch organic traffic grew from 95 to 1,650 sessions/month (+1,640%), direct booking +460%, Airbnb dependency from 88% to 48%, Muslim tourist mix from 8% to 32% — exceeding all targets.
The cooperative fund accumulated from the 18% revenue split reached THB 178,000 in the first 5 months, deployed on halal certification applications for multiple properties (paid and pending approval), shared marketing budget for the 2026 Hari Raya campaign, and a property maintenance fund for shared infrastructure (signage, parking, common area landscaping) — building a positive cycle Airbnb could never enable.
Lessons learned. Community-based tourism platforms must design revenue split + governance from day 1 — it’s not a feature to add later. The fact that Muslim tourism mix grew from 8% to 32% shows the Muslim Southeast Asia market is larger than Airbnb/Booking’s visibility suggests — whoever captures this niche first with infrastructure that respects religious requirements wins.
The client is currently in month 8 of retainer and expanding — they’re accepting applications from new homestays joining the network (there’s a waitlist) + starting phase 2 to add an activity marketplace where Trang local artisans can list their own workshops (batik dyeing, southern Thai cooking class, traditional medicine workshop). See SEO services emphasising local + multilingual, or other case studies.