Struck numbers = original Phase-1. The adversarial registry check found 5,258 "licensed massage BKK" = cumulative license records; active ≈ 2,522 — massage TAM re-based on the 3,174 actively-scraped shops. Salon/beauty rows unchanged; the −19 SOM sits entirely in the conditional Phase-2 segment.
DBD juristic registry, keyword-filtered to beauty-service objectives · Jan 2024 – May 2026 · click legend items to toggle series
StoreHub Salesforce leads (BigQuery th_ops_base) · Inbound cohort filter · sub-industry-tagged leads shown · quarterly
Ahrefs, Thailand, exact keyword, monthly volume
All 9 generic category terms combined stay flat at ~250/mo the entire 24 months
"sherwe" rising from zero (+175–200% YoY in Keyword Planner) = the LMWN clock ticking · "ease pos" fading
Off-book payments, supply theft, manager under-reporting. Local vendors literally lead with "ลดการทุจริต".
Day-job or investor owners hit a trust wall. Peer advice verbatim: "ต้องมีระบบที่ดี" — you need a good system.
Clients follow the person, and hunt them down via Facebook groups. Owning the client book is the defense.
Double-bookings, missed chats, conditions "buried in a long LINE message nobody reads".
Positioning consequences: frame pricing as "จ่ายปีละครั้ง" (pay once a year, never monthly), sell packages as "คอร์สที่ลูกค้าตรวจสอบได้" (verifiable balances), and attack GoWabi's ~20% commission + 45-day payout lag with direct payouts.
Packages/courses · Thai tax receipt + branch no. · loyalty/points. Sherwe includes courses in its base ฿12K tier; Ease and GoWabi POS ship all three.
POS register TH enablement (product live in JP since Jan 2026 - this is localization, not a build) · Opn QR acquiring (1.65% MDR; interim workaround already works: custom payment method + merchant's own PromptPay QR) · full inventory.
TH localization polish (฿25–80K) · ETDA e-Tax (optional in 2026; ~฿10/doc via API) · marketplace integrations · multi-branch dashboard.
| Gap | What's actually required (verified) | Cost / effort |
|---|---|---|
| Thai tax receipt + branch no. | Template with Revenue Code Sec 86/4 fields + 13-digit TIN + 5-digit branch code. Only VAT-registered merchants (>฿1.8M rev) issue tax invoices at all. | 1–2 wks · ~฿0 |
| PromptPay / QR acquiring | Opn: 1.65% QR / 3.65% cards, zero setup, KYC 3–15 days. Sell reconciliation + deposits — salons already get static QR free. | 2–4 wks |
| LINE reminders SHIPPED Apr 2026 | Booking confirmation + appointment reminder + cancellation notices released (LINE + email, configurable triggers). What remains is the economics design: LINE Basic = ฿15,360/yr ≈ the entire ARR → free reply messages + ≤1 push per booking, cost rides the merchant's own OA plan. | done · ฿0 |
| Localization + PDPA | ~10–25K words EN→TH + Thai line-break QA (JP i18n framework exists). PDPA: processor DPA + Thai notice + consent UX — enforcement is real (฿21.5M fines, Aug 2025). | 4–8 wks · ฿75–280K |
Total compliance/infra: ~฿100–300K + 7–15 eng-weeks (reminders came off the list — shipped). Every gap has an integrate-not-build path. The packages + loyalty engine is the only expensive item — and it doubles as the trust differentiator.
Jun–Aug eng capacity is committed to JP ICP needle-movers (auto-assignment · one-time messaging · customer imports · staff-app · coupon engine) under a stated "reduced-capacity reality". A Q3 gate commit must explicitly displace this list → Branch-B (Q3 foundation → Q4 launch) probability rises. The Aug coupon engine ≠ prepaid-package ledger — it does not close the packages P0.
Customer records are entered one at a time; JP already carries a churn-risk deal that was sold on manual import. Every TH switcher (Sherwe / Ease / paper) needs migration — switching cost is the gate on most deals. Soft launch = white-glove manual entry (fine at SOM scale); customer-master import targets Jul with a multi-format mapping layer — ask Product to add Thai incumbent export formats.
RELEASED: staff clock-in/out · customizable auto-notifications · digital karte. COMING: staff-app career levels + evaluation loop (Jul) · one-time messaging (Jul) · coupon engine (Aug). All reinforce the staff / ค่ามือ / anti-fraud entry. And JP's deepest objection — "retention tool, no NEW customers" (8/35 loss calls) — will hit harder in TH vs GoWabi: answer it with positioning + the Meta message-ads engine, not product asks.
| Feature (P0 = launch gate) | ||||||
|---|---|---|---|---|---|---|
| Packages / prepaid courses P0 | ❌ bundle menus ≠ prepaid ledger | ✅ | ✅ | ✅ | ✅ | ✅ |
| Thai tax invoice + branch no. P0 | ❌ receipt module live → ≈1-2 wks | ? | ✅ | ✅ | ◐ | ? |
| Loyalty / points P0 | ❌ stamp card = KIV backlog | ◐ PRO ฿24K | ✅ | ✅ | ◐ | ◐ |
| POS register (TH) | ◐ live JP Jan 26; TH = localize | ✅ | ✅ | ✅ | ✅ | ✅ |
| LINE confirmations & reminders | ✅ shipped Apr 26 | ✅ | ✅ | ◐ | ❌ | ◐ |
| Inventory | ◐ products | ✅ | ✅ | ❌ | ✅ | ✅ |
| LINE-native booking | ✅ | ✅ | ✅ Std+ | ◐ | ❌ | ❌ |
| Multi-channel booking (own site · Google · IG) | ✅ | ❌ LINE-only | ❌ | ◐ marketplace | ❌ | ◐ |
| 24/7 online booking | ✅ | ✅ | ✅ | ✅ | ◐ queue | ✅ |
| Commission / payroll (ค่ามือ) | ✅ | ✅ | ◐ | ◐ | ✅ | ◐ |
| Staff training / leveling | ✅ unique | ❌ | ❌ | ◐ | ❌ | ❌ |
| Multi-branch dashboard | ◐ multi-location core live | ◐ no unified view | ? | ? | ◐ | ✅ |
SalonX today: 6 ✅ · 3 ◐ · 3 ❌ (re-checked 3 Jul against the SalonX Phase-2 product tracker, 317 items: LINE reminders shipped Apr 2026 · POS/Register live since Jan 2026 · multi-location core live · merchants can add custom payment methods, so PromptPay-via-own-QR works as a recorded payment today). The three ❌ are exactly the three P0 gate items. Where SalonX is genuinely ahead: multi-channel booking (incumbents are LINE-only) and staff training/leveling (unique, and staff pain is trigger #1). Source: Jun-2026 verified competitor KB + Sales Dec'25 study; the SalonX column is corrected against the JP sales sheet — the Phase-1 doc over-stated loyalty and inventory. ✅ has · ◐ partial or tier-gated · ❌ missing · ? unverified.
The uncomfortable read: every SalonX gap is included at or below both price points by at least one credible competitor. Price is not the problem — feature parity at the price point is.
ร้านปิด… แต่ยังรับนัดลูกค้าได้ตลอด 24 ชั่วโมง — การันตีความไว้วางใจจากร้านค้ากว่า 300+ แห่ง ภายใน 1 ปี!
การันตีด้วยผู้ใช้กว่า 650+ ร้านทั่วประเทศ… เริ่มต้น 833 บาท/เดือน · Booking Website เชื่อมต่อ LINE OA เริ่มต้น 500 บาท/เดือน
ระบบจัดการการขาย (POS) ของโกวาบิ… เริ่มต้นเพียง 832 บาท/เดือน… เชื่อมต่อแอพพลิเคชันร้านค้าบน Gowabi
เพิ่มยอดได้ โดยไม่ต้องยิงแอด… ทดลองใช้ฟรี — LINE OA: @doctorease
Every player sells through chat (message ads or lead forms → LINE OA → free demo). There is no self-serve signup motion anywhere in the category. Logos are site favicons for identification.
10 months post-acquisition, no beauty booking in LINE MAN or Wongnai. JERA's integration page is still future-tense, clinics-only. Sherwe shows zero LMWN co-branding.
Beauty named a diversification pillar (Bangkok Post, Jun 2026) — and LY Corp's pre-IPO story (2 Jun 2026) explicitly lists beauty salons, spas, clinics in the 70k-merchant POS pillar. Cuts both ways: the 2027 IPO + ฿10B AI pivot compete for attention, but a declared diversification pillar can be accelerated to defend the equity story. Direction unknown — trust the tripwires, not the narrative.
Hidden unpublished "LMWN" page found in sherwepos.com's site code · Sherwe ads expanding into Pilates/Yoga · "sherwe" searches +175–200% YoY.
Integration queue is clinic-first → ICP-A salons come last. Monitor quarterly; exact tripwires below.
| Our tier | Lands against | Must include to be credible | Risk if missing |
|---|---|---|---|
| Entry — ฿12,000/yr | Sherwe LITE ฿12K head-on (LITE includes courses ⚠) · ThaiHand ฿10K just below · POSPOS/Qashier budget floor ฿2.5–8K | Booking + POS + LINE reminders + packages | Entry without packages = DOA vs LITE — the ฿12K shopper compares line-by-line |
| Standard — ฿18,000/yr | Sherwe STANDARD ฿18K · equals StoreHub TH salons' observed weighted-avg ARR ฿18K (58 wins) | Entry + loyalty + multi-channel booking + marketing automation + staff app | Parity gap until P0 closes; multi-channel + staff features are the over-deliver |
| Premium — ฿24,000/yr (later, not launch) | Sherwe PRO ฿24K (loyalty gated there) · SilomPOS top ฿29.9K | Multi-branch dashboard + advanced analytics + priority support | Premium without a mature multi-branch view won't clear PRO — park it |
| Structure | Blended ARPU | Conservative 25 won | Standard 42 won | Optimistic 63 won | Paid-CAC payback* | Verdict |
|---|---|---|---|---|---|---|
| A · single ฿12K | ฿12,000 | RM 39,735 | RM 66,772 | RM 100,159 | ~17 mo ⚠ | Fastest door-opener vs the budget floor — weakest unit economics; paid CAC can exceed year-1 ARR |
| B · single ฿18K | ฿18,000 | RM 59,603 | RM 100,132 | RM 150,199 | ~11 mo | Best unit economics — but P0-hostage: must beat Sherwe STANDARD on feature parity from day 1 |
| C ★ two-tier ฿12K entry + ฿18K standard recommended | ≈฿15,000 50/50 mix | RM 49,669 | RM 83,444 | RM 125,166 | ~14 mo | Entry price kills the price objection at the door; standard tier captures the validated ฿18K WTP — packages at entry, loyalty + multi-channel at standard |
How to read this honestly: win counts are held identical across A/B/C (the Q5 funnel model) because there is no TH salon price-elasticity data — the direction sits in the verdicts, not the math. Structure C's 50/50 mix is anchored to StoreHub TH's own salon base: 27 of 58 wins (47%) sit on the ฿12.5K starter tier, yet weighted-avg ARR = ฿18K — both price points demonstrably transact. Revenue cells = wins × price ÷ ฿7.55/RM (Jul-2026; earlier cells mixed three implied rates — fixed 4 Jul); multi-branch bump excluded (adds ~+6–9%). Realized ARPU on the historic salon base is ฿14.1K avg / ฿12.5K median — treat C's ฿15K as mildly optimistic. *Payback = media-only CAC ÷ ARR × 12 at the model point (฿600 CPL ÷ 3.5% = ฿17.1K); halves at ฿300 CPL. Renewal-adjusted: at the observed 59–65% Y1 renewal (→ Investor lens), ~35–40% of a paid cohort never completes the A/C paybacks — every payback figure assumes the merchant renews once.
The cliff sits exactly at the renewal decision: 85% survive past month 12, only 59% past month 15. Of 73 all-time beauty wins, ~34% are active subscribers today. This corrects the earlier draft's "renewals improve the picture" line — they add LTV, but they do not rescue paid CAC. One genuinely new lever: a TH renewal-opportunity pipeline went live Dec-2025 (742 opps · 7 won · 685 open) — true renewal-rate cohorts read out within a quarter (added to monitors).
| CAC ceiling for a 3:1 business | ≈ ฿11–13K revenue-basis · ฿8–10K on gross margin |
| Media-only CAC @ model point (฿600 CPL ÷ 3.5%) | ฿17.1K → ~2:1 revenue-basis — already fails |
| Fully-loaded CAC at Y1 volumes (+ rep + CX labor ÷ 25–42 wins) | ฿42–59K → <1:1 — scale dilutes labor, but not inside Y1 |
Paid-alone acquisition fails the 3:1 bar at ANY plausible salon CPL. The economics only work where the doc already points: zero-CAC channels first, renewal lift, and volume that spreads the fixed team.
| 100 leads/mo | ~22 demos booked (22.3% lead→SQL, our funnel) → ~11–14 held (45–68% show) → 3.5 won |
| Sales capacity | 1 full-cycle rep caps at ~80–100 leads/mo; 42 wins/yr = a full SMB-AE annual quota |
| Onboarding wall | No bulk import → white-glove migration ≈ 2–7 CX-days per switcher → ~4 onboards/mo per CX head (standard case saturates one) |
| Minimum team | 2 dedicated Thai heads ≈ ฿1.0–1.4M/yr loaded — ~3× conservative Y1 bookings |
| Line | Band (THB) | Basis |
|---|---|---|
| One-time build | ฿0.73–1.53M | Compliance ฿100–300K + 7–15 eng-wks (verified) + first sizing of the P0 packages+loyalty build: 14–26 eng-wks @ ~฿30K/loaded-wk (BKK senior eng ฿100K/mo, Michael Page) — analyst estimate, needs eng sign-off |
| GTM team | ฿0.72–1.38M /yr | 1 sales/chat rep + 0.5–1 CX/onboarding, Bangkok bands (JobsDB/Glassdoor, verified) |
| Media | ฿0.40–0.74M /yr | Steady-state paid to hit 60–100 leads/mo at ฿700 mid-CPL (organic supplies only ~12/mo); 90-day instrumentation tranche ฿60–300K, point ask ~฿150K ≈ 200 leads |
| Total Y1 cash | ฿1.9–3.8M (central ~฿2.9M ≈ RM380K) | vs Y1 bookings ฿0.38–0.95M → net Y1 burn ≈ ฿1–3.4M. This is a multi-year bet, not a Y1-payback project |
Breakeven: on a lean stack (shared 0.5 rep + 0.5 CX, zero-CAC-led, ~฿570K/yr opex) cumulative breakeven lands around month 29 (B ฿18K) · 34 (C) · 48 (A) at an assumed 70% renewal — the observed 59–65% pushes those later. On a fully-dedicated stack (~฿1.5–1.9M/yr opex) the BKK-salon SOM revenue ceiling (102 × ฿15–18K ≈ ฿1.5–1.8M) roughly equals opex — inside BKK salons alone it never breaks even. The case therefore requires: shared/fractional headcount with StoreHub TH · T1-metro expansion (+~30% SOM → ~133) · massage Phase-2 · payments margin. All four are plannable; none is currently committed.
LMWN's proven template: Wongnai POS runs −11 to −31% promos + a free LINE Pay QR box. Needs no board decision, cross-subsidized by food-delivery GP. Our pre-committed answer (monitors): hold price, reframe on multi-channel + staff + no-marketplace-conflict — never chase below ฿10K.
POS already ฿1,040–1,250/mo with marketplace promotion bundled; could go ~free for commission-paying venues to protect the ~20% GMV take. But total funding is ~$4M (last raise Aug 2022) — can start a bundling fight, cannot fund a long subsidy war.
TH is not on Fresha's card-processing country list and its only BKK role is portfolio-farming (1 AM hire). PromptPay's zero-MDR rail weakens its free-software + processing model here. Watch quarterly; not a launch-window threat.
Platform-dependency risk (unresolved): the entire funnel (Meta message ads → LINE OA, the LINE Mini-App wedge) rides LY Corp rails — the parent of the #1 competitor. Plan B belongs in GTM prep: own web booking page, Messenger/phone fallback, exportable customer data. Also note the beauty clock runs regardless of our entry — LMWN's pre-IPO story already names the vertical; a ~102-SOM entrant likely doesn't reprioritize a ฿10B agenda, but a price response needs no reprioritization.
The mechanism exists: Opn "Account Chaining" (beta, TH-documented, by-request) and GBPrimePay — now Xendit — has a platform-fee product (TH terms unverified). At a 0.2–0.5% ISV spread on 25–50% attach, a ฿1.2–2.4M-GMV salon yields ~฿0.9–4.5K/merchant/yr ≈ 6–30% of blended ARPU. Structural cap: 39.5% of salon payments ride zero-MDR PromptPay — routing it through a 1.65% rail is a reconciliation sell, not a default. StoreHub TH's book today is software-only (0 Beep wins 2026 YTD). Cheap next step: one BD call with Opn for actual rev-share terms.
Renewal-adjusted 24-month cumulative (Structure C, 65% renewal): RM82K / 138K / 207K; Y2 exit run-rate ARR ≈ RM81–137K — the optimistic path breaches BKK SOM (126 cumulative wins > 102) by mid-Y2. Growth after ~month 18 must come from: T1 metros (+~30% SOM → ~133; other-T1 = 12.7% of DBD inflow) · massage Phase-2 · multi-branch upsell (+6–9% ARR; 17.0% of the base is multi-branch, verified). Install-base mining is small: 52 active beauty-service accounts (replacement revenue) + 36 beauty-retail (referral pool only).
Still open after this pass (flagged, not answerable from here): gross margin per merchant at white-glove intensity (needs Finance COGS actuals) · SalonX JP cohort retention (best predictor of TH churn — needs JP subscription data) · eng sign-off on the 14–26 eng-wk P0 estimate · Opn/Xendit actual rev-share terms (BD call).
| Scenario | Leads /mo | Won Y1 | ARR @ ฿12K | ARR @ ฿18K |
|---|---|---|---|---|
| Conservative | 60 | 25 | RM 39,735 | RM 59,603 |
| Standard | 100 | 42 | RM 66,772 | RM 100,132 |
| Optimistic | 150 | 63 | RM 100,159 | RM 150,199 |
A fair comparison is salon-scoped CPL vs salon-scoped CPL. JP's RM896–1,299 is a salon-vertical number; StoreHub TH's proven RM50–70 is broad retail/F&B — it proves the funnel mechanics, not the salon price. Scoped to salon/beauty, the TH audience is far smaller, so cost per qualified lead will carry a niche premium over the broad benchmark. TH salon CPL is unvalidated in-market, but it is now modeled from measured inputs (below): band ฿450–950/lead, plan at ≈฿650 — a tightening of the earlier ฿400–1,000 assumption, not a replacement. The structural TH advantages that still hold: base click costs ~5–10× below JP, a free LINE chat funnel, and hard evidence the vertical's unit economics work — ThaiHand sustains ~28 always-on ads selling ฿10K/yr software and built its brand search 12× doing it. That is impossible at JP-level CPLs against these ARRs.
| Input | Value | Source |
|---|---|---|
| TH broad cost / messaging conversation | RM33.5 ≈ ฿253 | Measured: FB/TH Engagement-Message campaign, RM20.5K spend → 612 conversations, last 90d |
| Conversation → qualified SF lead | ~55–65% | Implied: ฿253/conversation vs the proven ฿390–550/SF-lead on the same funnel |
| Audience narrowing (TH, 20–55) | 49.3M → 14.2M | Meta delivery estimates: broad → beauty-consumer stack 25.3M → cosmetology/pro stack 14.2M. Key: even the "tight" stack is 14M people — interest targeting cannot isolate ~40K salon operators, so the cost sits in creative self-selection, not CPM |
| Niche multiplier | ×1.3–2.5 | Modest CPM premium (auction liquidity fine at 14M) + owner-self-selection waste vs the broad merchant ads; bounded below by vertical-creative relevance, above by category evidence |
| Category sustainability bound | ≤ ~฿700 | ThaiHand sustains ~28 always-on ads selling ฿10K/yr software (2+ yrs, brand search 12×) — only operable if vertical CPL sits in the mid-hundreds ฿ at ~7% conversion |
| JP reality check | RM828 / CPM RM130 | Measured on our JP SalonX leads campaign (90d). TH impressions cost ~2–3× less (CPM RM45 vs 130) and the chat funnel converts where forms do not — the structural reason TH ≠ JP |
| Result — salon/beauty | ฿450–950 · plan ≈฿650 (RM86) | = ฿253 × (1.3–2.5) ÷ (0.55–0.65). Spa/massage: same band, likely the low end (larger owner pool; ThaiHand is the massage-side proof). Barbershop: treat as salon |
Implications, unchanged in direction: the ฿150K instrumentation tranche buys ~230 leads at the ฿650 central · the kill-criterion cap stays ฿1,000 · at ฿650 CPL and 3.5% conversion, media-only CAC ≈ ฿18.6K — still above the ฿11–13K LTV ceiling, so the zero-CAC-first sequencing stands. The model is an estimate stack, not market data — the first 90 days replace it with real numbers.
| CPL ↓ · Conversion → | 3.5% (model) | 7.2% (qualified organic) |
|---|---|---|
| ฿300 /lead | ฿8.6K · 71% A / 48% B | ฿4.2K · 35% / 23% |
| ฿600 /lead | ฿17.1K · 143% A ⚠ / 95% B | ฿8.3K · 69% / 46% |
| ฿1,000 /lead | ฿28.6K · 238% ⚠ / 159% ⚠ | ฿13.9K · 116% ⚠ / 77% |
The honest read: at plausible salon-scoped CPLs, paid-only CAC can exceed year-1 ARR at model conversion. Two caveats cut both ways: the 7.2% is organic, self-selected leads — paid usually converts below organic; and renewals are now measured, not assumed: 59–65% Y1 renewal on our own base — enough to add ~1.6 renewal-years of LTV, not enough to rescue paid-only CAC. Also note the optimistic case (63 won) = 62% of BKK SOM captured in year 1 with zero competitive haircut, while every conversion input comes from uncontested organic inbound — post-launch, each paid lead demos 2–3 chat-based rivals. This is why the plan front-loads the zero-CAC plays and treats the CPL→SQL→won chain as the launch's go/no-go instrument.
| Leads/mo ↓ · Conv → | 2% | 3.5% (model) | 7% (qualified-lead ceiling) |
|---|---|---|---|
| 60 | 14 | 25 | 50 |
| 100 | 24 | 42 | 84 |
| 150 | 36 | 63 | 126 |
The 7% column is a ceiling, not a base case — it matches organic salon conversion; paid traffic typically lands below it. Zero-CAC channels (install base, GoWabi list, distributors, academies) convert nearer the organic rate, which is exactly why they lead the sequencing.
| LMWN beauty go-live | jeracloud.com/lmwn flips to live-feature copy · hidden Sherwe "LMWN" page publishes · sherwe brand-search jump | Quarterly |
| Fresha escalation | More Bangkok hires · TH payment rails appear · venues > 200 | Quarterly |
| Price floor | GoWabi < ฿832/mo · ThaiHand < ฿10K · Sherwe LITE discounting — pre-committed answer: hold price, reframe (multi-channel + staff + no-marketplace-conflict); never chase below ฿10K | Monthly |
| TH renewal reads | Renewal-opp pipeline live since Dec-25 (742 opps · 7 won · 685 open) — first true renewal cohorts read out in Q3; feeds the LTV model | Monthly |