LeakStop — payout auditor for small tour operators

1. One-liner

Catches every booking the OTAs underpaid or never paid, and hands you the recovery line item.

2. Trend signal — why now?

Tours-and-activities is a $179B market in 2024 heading to $264B by 2030 (6.7% CAGR), and 72–94% of operators are micro/small businesses — a kayak guide, a cooking-class host, a family van-tour outfit. They sell across Viator, GetYourGuide, Klook and their own site, each with its own commission rate (Viator ~20%, GYG ~30%), its own payout cycle (Viator pays ~21 business days after the travel month; GYG monthly by default, bi-weekly for an extra 2%), and its own statement format.

The result: a documented reconciliation nightmare. An industry write-up puts it bluntly — tracking “what Viator owes you, what GYG owes you, what direct bookings already settled… without a spreadsheet that becomes a full-time job by July.” A commission-tracking firm (Sion) estimates more than 40% of travel commissions have discrepancies that need reconciling or never get paid, and 20%+ get lost in the shuffle entirely. Operators report spending up to 20 hours a week, or two full days a month, matching bookings to bank transfers in spreadsheets — where one mistyped digit cascades into wrong cash-flow forecasts.

What changed in the last 12 months: (1) Viator’s Supplier API and downloadable commission/booking reports matured, GYG shipped a new commission-breakdown system, so the raw data is finally exportable; (2) cheap LLM document parsing makes ingesting messy PDF/CSV payout statements a weekend job instead of a quarter of OCR engineering; (3) vertical-SaaS investors are openly chasing embedded-finance and reconciliation plays for the world’s MSMEs.

Nobody owns the inbound payout-audit lane for small operators. Channel managers (Bokun, Rezdy, FareHarbor, Checkfront) sync availability, not money. Automate.travel and full booking-ops platforms fold finance into a heavy all-in-one. The narrow, cheap “did the OTA actually pay me what it owed, per booking?” tool doesn’t exist for this wallet.

Provenance:

• Signal 1 (demand): “Sion estimates more than 40% of commissions either have discrepancies that need to be reconciled or do not get paid… more than 20% get lost in the shuffle and go unpaid”; agents spend up to 20 hrs/week reconciling — https://www.travelweekly.com/Travel-News/Travel-Agent-Issues/New-solutions-introduced-collect-commissions — 2026-06-02
• Signal 2 (feasibility): Viator Supplier/Partner API with downloadable booking & commission reports; GYG new commission-breakdown system; reconciliation “becomes a full-time job by July” without tooling — https://docs.viator.com/supplier-api/technical/ and https://automate.travel/blog/viator-vs-getyourguide-for-operators/ — 2026-06-02
• Signal 3 (economic): Tours & Activities reservations market USD 179B (2024) → 264.4B (2030) at 6.7% CAGR; 72–94% of operators are micro/small businesses; 20–30% OTA commissions on the line — https://www.researchandmarkets.com/reports/5310766/tours-and-activities-reservations-global and https://www.sambahq.com/ota-supplier-guide/ota-commission-rates — 2026-06-02 Category: Workflow automation

3. The opportunity

The pain isn’t “I can’t see my bookings.” It’s “I have no idea if the money landing in my bank account is correct.” Every OTA payout is a black box: a lump sum, weeks late, against bookings that travelled across a fuzzy date range, net of a commission that may or may not match the contracted rate, minus refunds and cancellations the operator may not have tracked. When 40% of commissions carry discrepancies, a small operator doing 1,000 bookings/year on a $100 average ticket is plausibly leaking four-to-five figures a year in short-pays, missing payouts, and over-deducted commission — and they can’t prove it because the audit work is too tedious to do by hand.

The incumbents don’t touch this. Channel managers solved the front of the funnel (don’t double-book). The back of the funnel — reconciling cash actually received against cash owed — is still a spreadsheet. We do one thing the all-in-one platforms bury and the channel managers ignore: ingest every OTA statement and every booking record, match them line by line, and surface exactly which bookings were underpaid, unpaid, or over-commissioned — with a recovery dossier the operator can paste into a supplier-support ticket. It pays for itself the first month it claws back one missing payout.

4. Target market

• Primary customer: Owner-operators and ops managers of small tours-and-activities businesses doing roughly 300–5,000 bookings/year across 2+ OTAs plus direct — day-tour outfits, dive shops, food-tour hosts, adventure operators, boat charters. Solo or up to ~10 staff. Global, with a deliberate tilt toward price-sensitive SEA and LatAm operators where a $39/mo tool sells but a $300/mo ops suite doesn’t.
• Why they buy: “I’m fairly sure Viator and GetYourGuide aren’t paying me everything they owe, but I’d have to spend two days a month with a spreadsheet to prove it, and I’d rather run tours.” The money is real and recoverable; the audit labor is what’s missing.
• Rough TAM reasoning: Hundreds of thousands of small operators globally are listed on the major OTAs. Even capturing a few thousand at $39–99/mo is a healthy bootstrapped business well inside the sub-$5M ARR target.
• Why now for them: OTA commissions are rising (GYG pushed increases past 30% for some in mid-2025), so every leaked dollar hurts more, and the statement data is finally exportable via API.

5. Product sketch (MVP)

• Connect Viator, GetYourGuide, Klook and a direct-booking source (Stripe/booking engine) once.
• Auto-ingest every payout statement (PDF/CSV) and every booking record — LLM-parsed, no manual entry.
• Line-by-line match: each booking → expected net payout (price − contracted commission − tracked refunds) vs. actual amount received.
• A leak report: bookings underpaid, never paid, or over-commissioned, with dollar totals and aging.
• One-click recovery dossier per discrepancy — booking ref, expected vs. actual, statement screenshot — formatted to paste into Viator/GYG supplier support.
• Cash-flow view: “you’re owed $X across Y bookings; $Z is overdue past the platform’s own payout SLA.”
• Monthly “you recovered $X with LeakStop” summary — the renewal-justifying receipt.

6. AI angle — what’s load-bearing

AI does two things that make this buildable by a tiny team. First, statement ingestion: OTA payout statements come as inconsistent PDFs and CSVs with shifting columns, bundled refunds, and currency conversions — an LLM extraction layer turns that mess into structured line items without per-platform OCR engineering. Second, fuzzy matching and explanation: bookings and payouts rarely align one-to-one (date ranges, partial refunds, FX, batched lump sums), so the model reconciles many-to-many and, critically, explains in plain language why a line is short (“paid as if commission were 30% but your contract says 25% — $4.20 under”). Remove the AI and you’re back to the two-day spreadsheet this product exists to kill. The matching judgment is the product.

7. Localization angle (if any)

Real wedge, not forced. The same operators in Bali, Cusco, Hanoi, or Cape Town feel this harder — OTA payouts are their lifeline, FX conversion adds another layer of leak, and global ops suites are priced for US/EU wallets. A ₹/Rp/local-tier at a fraction of the all-in-one price, with multi-currency reconciliation and FX-spread flagging built in, beats a generic global tool. Payout SLAs and commission norms are universal across OTAs, so the engine ports cleanly; only pricing and currency handling localize. This also pulls the portfolio away from its recent US concentration.

8. Business model — path to $1M–$5M ARR

• Pricing: $39/mo (SEA/LatAm/solo tier) to $99/mo (multi-OTA, higher volume). Optional success fee — 10–15% of recovered dollars — for operators who’d rather pay on results.
• ACV: ~$700–1,000/yr blended on subscription; meaningfully higher where the success fee kicks in on a big recovered backlog.
• Rough math to $1M ARR: ~1,200 operators × ~$70/mo avg × 12 ≈ $1.0M. Inside the small-operator universe on the major OTAs, very reachable.
• Rough math to $5M ARR: ~5,000–6,000 operators plus success-fee revenue on recovered backlogs, plus a per-OTA add-on as we cover more platforms (Klook, Musement, Tiqets, GetTransfer). Requires becoming the default audit layer in 2–3 operator communities.
• Expansion path: more connected OTAs, success-fee recoveries, then adjacent reconciliation (payment-processor fees, refund-leakage, tax-withholding checks) — each a new line item on the same data.

9. Go-to-market wedge — first 100 customers

• Scrape the OTAs themselves. Viator/GYG/Klook list operators publicly with contactable websites. Pull 2,000 small operators running on 2+ platforms, send a personalized teardown — “we sampled your last 3 months of public booking volume; operators your size typically have $X in OTA payout leakage” — and offer a free first audit. A free audit that finds real money converts; expect low-single-digit % to paid, enough to seed 100.
• Operator Facebook groups and forums. Tour-operator communities, Arival, and OTA-supplier groups are where commission gripes already happen (GYG’s 2025 rate hike lit them up). Show up with a “we found $4,200 GYG never paid this operator” case study, not a pitch.
• Free “OTA Payout Leak Check.” A landing page where an operator uploads one Viator/GYG statement and gets an instant discrepancy estimate — lead magnet that demonstrates value in 60 seconds and feeds the paid funnel.
• Bookkeeper / virtual-assistant partnerships. The people doing the two-days-a-month reconciliation are often outsourced VAs and travel bookkeepers — partner with them; we make their deliverable instant and they resell us.

10. Build complexity — justification

Medium. Off-the-shelf: web stack, Stripe/booking-engine connectors, LLM statement parsing, Viator’s documented Supplier API. The custom work is the reconciliation/matching engine and per-OTA statement adapters (each platform’s format and quirks), plus keeping connectors alive as OTAs change formats. A technical founder with a travel-ops advisor ships a credible v1 covering Viator + GYG + direct in ~3–4 months; Klook and others follow as adapters.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Reading an operator’s own payout data on their behalf; no scraping of OTA private areas without operator consent/credentials.
Ethical — no harm / dark patterns	✅	Helps small operators recover money they’re owed. The only “loser” is an OTA that underpaid.
Market exists (evidence above)	✅	40%+ commission-discrepancy stat, documented 20-hrs/week reconciliation pain, rising commissions.
1–5 person team can build this	✅	Technical founder + advisor, off-the-shelf APIs + LLM parsing.
Launchable with <$50K / ₹40L	✅	Solo/pair build, no inventory, no regulatory spend.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Real recoverable money + tedious 2-day/month labor. Not hair-on-fire daily, but felt every payout cycle and quantifiable in dollars.
Demand evidence	15	12/15	Strong: 40% discrepancy stat, rising commissions, public “full-time job” framing. Docked for thin verbatim operator quotes (Reddit inaccessible; forums sparse).
Build feasibility	15	11/15	Documented APIs + LLM parsing, but per-OTA adapters and a real matching engine push it past a weekend.
Distribution clarity	15	11/15	Operators are publicly listed and scrapeable; free-audit lead magnet is concrete. Conversion on cold outreach is the unknown.
Revenue mechanics	15	11/15	Pricing benchmarks against existing ops tools; success fee adds upside. ACV modest, so volume must come.
Time to first revenue	10	8/10	Free-audit-to-paid funnel can close in weeks once 1–2 OTA adapters work.
Defensibility	10	6/10	Soft moat: accumulating per-OTA statement-format knowledge, recovered-dollar track record, operator trust. Copyable, but a head start in the audit niche compounds.
Total	100	75/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required

Key assumptions to validate (3–5)

1. Assumption: Real, recoverable leakage exists at the per-operator level (not just industry-aggregate 40%). How to test: Hand-audit 10 operators’ last 3 months of Viator/GYG statements; if median recoverable < ~$200/mo, the value prop is too thin.
2. Assumption: Operators will connect their OTA accounts / upload statements to a new tool. How to test: Offer 30 operators a free manual audit in exchange for statement access; measure how many say yes.
3. Assumption: Statements are parseable reliably enough to trust the leak report. How to test: Run LLM extraction on 50 real statements across formats; measure extraction error rate before promising “we caught it.”
4. Assumption: $39–99/mo clears for this wallet, especially SEA/LatAm. How to test: Price-test in operator FB groups and on the free-audit follow-up.

Risk flags

1. Platform dependency: Built on OTA statement formats and APIs. If Viator/GYG change formats or restrict API access, adapters break. Mitigate with statement-upload fallback (operator-supplied data, no API dependency).
2. OTA retaliation / ToS friction: An OTA may dislike a tool that systematically disputes its payouts. Mitigate by framing as the operator’s own audit of their own data — operator-initiated recovery tickets, not third-party intervention.
3. Value ceiling per customer: If average recoverable leakage is small for the smallest operators, ACV stays low and you need real volume. Mitigate with the success-fee tier and moving upmarket to higher-volume operators.

14. Structured verdict

Score:                  75/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder + travel-ops/bookkeeping advisor
Time to revenue:        6–10 weeks after first 2 OTA adapters work
Capital to launch:      $5–10K / ₹4–8L
Top 3 assumptions to validate first:
  1. Per-operator recoverable leakage ≥ ~$200/mo median — hand-audit 10 operators' 3-month statements
  2. Operators will grant statement/API access — offer 30 free manual audits, measure yes-rate
  3. Statement parsing is reliable — LLM-extract 50 real statements, measure error rate
Kill criteria:
  - Abandon if median recoverable leakage across 10 hand-audits is under ~$200/mo (value too thin to price)
  - Abandon if <15% of 30 operators offered a free audit will grant statement access (distribution dead)

15. Next step — 1-week validation sprint

• Day 1–2: Recruit 10 small operators from tour-operator FB groups; collect 3 months of Viator + GYG payout statements and booking exports under a “free audit” offer.
• Day 3–4: Manually (with LLM assist) reconcile their statements against bookings. Quantify recoverable leakage per operator and note how messy each format is.
• Day 5: Decide go / no-go on a falsifiable bar — median recoverable leakage ≥ ~$200/mo AND ≥6 of 10 operators react with “yes, find me that money, here’s my account.” Miss either and the idea is too thin to price; hit both and the free-audit funnel is the whole GTM.