SlotSentry — overbooking tripwire for small tour operators

1. One-liner

SlotSentry watches every OTA you sell on and pings your WhatsApp the second two channels are about to oversell the same boat.

2. Trend signal — why now?

The tours-and-activities long tail still runs on spreadsheets, WhatsApp groups, and three separate OTA dashboards. The numbers are blunt: 39% of tour operators worldwide have no booking system at all, and among small operators (under 1,000 guests/year) only 42% have adopted any system (Arival State of Booking Tech, via automate.travel). The industry moves $271B/year through manual entry. Overbooking is the recurring scar tissue: sell your last two seats on Viator and GetYourGuide still shows them available 2–5 minutes later, and you’ve double-sold a 12-seat boat.

This is not theoretical. A captain in Naxos counted 13 guests for a 12-seat boat — two booked the day before on Viator, one an hour earlier on GetYourGuide — because the €49/mo channel manager he paid for synced on a 22-minute delay. Result: three angry guests, two refunds, a TripAdvisor rating that dropped from 4.8 to 4.6 (captainbook.io). A RIB-safari operator selling on GetYourGuide, Viator, his own site, and two local hotels spends two hours every morning manually re-entering bookings into a spreadsheet to avoid exactly this.

Why now: GetYourGuide’s API still has a 2–5 minute availability propagation delay, Viator’s Partner API is gated behind an apply-and-certify process (docs.viator.com), and the full channel managers (Peek Pro, FareHarbor) charge 6–8% per booking (bokun.io pricing) — so the smallest operators stay on spreadsheets. WhatsApp Business API + cheap LLM email parsing now make a lightweight, no-migration alarm buildable for the first time.

Provenance:
  - Signal 1 (demand): 39% of tour operators have no booking system; only 42% of small operators have adopted one; verbatim overbooking horror stories (Naxos 13-for-12 boat, RIB operator's 2hr/morning spreadsheet) — https://automate.travel/blog/tour-operator-software-guide/ , https://www.captainbook.io/blog/how-to-prevent-double-bookings-across-ota-channels-(2026) — 2026-06-05
  - Signal 2 (feasibility): OTAs send per-agency booking confirmation emails operators already auto-forward/parse via Zapier; Viator/GYG Partner APIs exist but are gated; WhatsApp Business API + LLM parsing mature — https://docs.viator.com/partner-api/ , https://community.zapier.com/how-do-i-3/automating-bookings-from-online-agencies-31962 — 2026-06-05
  - Signal 3 (economic): Peek Pro/FareHarbor charge 6–8% per booking; Bokun 1–1.5%; funded category (Tripadvisor owns Bokun, Peek funded) on $271B GMV — https://www.bokun.io/peek-pro-pricing — 2026-06-05
  Category: Workflow automation

3. The opportunity

The channel-manager category solves sync — if you migrate onto it. That’s the catch. The full platforms (Peek Pro, FareHarbor, Rezdy, Bokun) ask a one-boat operator to move his whole business onto their CRM and pay 6–8% per booking. The data says the long tail refuses: 58% of small operators run no system. They keep selling on Viator + GetYourGuide + their own site + WhatsApp, and they eat the occasional double-booking as a cost of doing business — until the TripAdvisor hit costs them a season.

SlotSentry doesn’t ask them to migrate. It sits on top of the channels they already run, ingests every booking the moment it lands (forwarded confirmation email or a gated API where they have it), maintains the one unified live count nobody else has, and trips a WhatsApp alarm before they oversell — not 22 minutes later. It’s the smoke detector, not the new house. The incumbent can’t sell this because their entire model is “move onto our platform”; a tripwire that keeps you on the OTAs is anti-thetical to a channel manager’s lock-in pitch.

4. Target market

• Primary customer: Owner-operators of small tour/activity businesses — boat tours, RIB safaris, walking tours, diving, food tours, day trips — running 2–8 capacity-constrained departures/day, selling on 2–4 OTAs plus their own site, with 0–5 staff and no dedicated reservations person. Concentrated in Mediterranean (Greece, Croatia, Portugal, Spain), LATAM (Mexico, Colombia, Peru), and SEA (Indonesia, Thailand, Philippines).
• Why they buy (their words): “Sell your last 2 spots on Viator and GetYourGuide still shows them available — double-booking and an angry phone call at 7 AM.” “I block availability just in case and lose real sales.” “Two hours every morning updating a spreadsheet.”
• Rough TAM: Hundreds of thousands of tour/activity suppliers list on Viator (300K+ products) and GetYourGuide. Even the unautomated slice — the 39–58% on no system — is six figures of operators. Capturing 5,000 of them at ~$30/mo is a $1.8M ARR business.
• Why now for them: OTA volume keeps rising, propagation delays persist, and one bad-rating overbooking now costs more than ever because rating is the OTA ranking signal.

5. Product sketch (MVP)

• Forward-to-onboard: operator sets a forwarding rule (or forwards manually) for Viator/GetYourGuide/Klook/website booking confirmation emails to a unique SlotSentry address. No API approval needed to start.
• Unified live count per departure: one screen showing true remaining capacity per tour/time-slot across every channel + manually-added WhatsApp/phone bookings.
• The tripwire: the instant ingested bookings across channels approach or exceed a slot’s capacity, fire a WhatsApp alarm (“⚠️ Sunset Cruise 18:00 today: 12/12 sold across Viator+GYG — pause GYG now”) with a one-tap link to the channel to close.
• Manual booking capture: quick WhatsApp/web entry for the walk-up or phone booking so the count stays honest.
• Daily morning manifest: one WhatsApp message at dawn listing every departure, who’s coming from which channel, and any slot at risk — replaces the 2-hour spreadsheet.
• Oversell post-mortem: when an oversell does slip through, a clean record of which channel sold the extra seat and when — for the refund/rebook decision.
• Optional gated-API upgrade: for operators who already hold Viator/GYG Partner API access, connect directly for near-real-time instead of email.
• Multilingual (Spanish, Portuguese, Greek, Bahasa) WhatsApp output.

6. AI angle — what’s load-bearing

Two places AI does real work, not decoration:

1. Email parsing across messy per-OTA templates. Every agency (Viator, GetYourGuide, Klook, the operator’s own Stripe/site) sends a differently-formatted confirmation, in different languages, and they change templates. A brittle regex parser breaks weekly. An LLM extractor that reliably pulls tour name, date, time-slot, pax count, channel from any unseen template — including a screenshot a panicked operator pastes — is the core. Remove it and you’re back to per-template Zapier hell that operators already abandoned.
2. Slot matching / fuzzy reconciliation. “Sunset Sailing 6 PM” on Viator vs “Coucher de soleil — 18h00” on GYG vs “sunset trip” the operator typed into WhatsApp are the same departure. AI resolves these to one capacity bucket so the count is trustworthy. Without it, the unified count is garbage and the alarm cries wolf.

7. Localization angle

This is a localization-led play, not a global-generic one. The incumbents are US/EU-platform-first and priced in booking fees that hurt most in lower-AOV markets. The wedge markets are Mediterranean + LATAM + SEA where: (a) operators live in WhatsApp, not email or a web CRM — so a WhatsApp-first alarm beats a dashboard nobody opens; (b) tours are multilingual and OTA confirmations arrive in local languages — the AI parser must handle Spanish/Portuguese/Greek/Bahasa; (c) a flat $19–39/mo tier works where 6–8%-per-booking does not; (d) distribution runs through vernacular operator Facebook/WhatsApp groups the global vendors don’t touch. A ₹/peso/euro-friendly flat price + WhatsApp-native delivery is the localized cut that beats a generic channel manager for this segment.

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

• Pricing: flat $29/mo Solo (up to ~3 channels, 1 location), $59/mo Pro (unlimited channels, multi-boat/multi-guide, daily manifest, post-mortems). No per-booking fee — the explicit anti-Peek-Pro positioning.
• ACV: ~$420/yr blended.
• To $1M ARR: ~2,400 operators at $35/mo blended × 12. The unautomated long tail alone is far larger than 2,400.
• To $5M ARR: ~12,000 operators, or 6,000 + a higher-value add-on (auto-pause via API, payment-deposit chase, review-recovery). Reaching 12K means expanding from boats to all capacity-constrained activities (escape rooms, classes, rentals) — same overbooking shape.
• Expansion path: Solo→Pro on multi-departure growth; usage add-ons (auto-pause, deposit capture); a thin per-location fee as operators add boats/tours.

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

• Mine 1-star OTA/TripAdvisor reviews mentioning “overbooked,” “double booked,” “waited and there was no room.” Each names an operator. DM the operator (not the reviewer): “Saw the overbooking review — here’s a 60-second Loom of an alarm that would’ve caught it.” Expect a high reply rate; this is their fresh wound.
• Post a tear-down in the operator-heavy Facebook groups (“Tour & Activity Operators,” regional captain/guide groups in Greece, Mexico, Bali): the Naxos 13-for-12 story + “here’s the free morning-manifest WhatsApp.” Convert with a free Solo tier capped at one departure type.
• Partner with regional OTA onboarding consultants / local DMC WhatsApp communities who already advise small operators on Viator/GYG setup — they have the list and the trust; rev-share.
• Free “morning manifest” loss-leader: the dawn WhatsApp manifest is genuinely useful even before the alarm matters; give it away to seed the unified count, upsell the tripwire.
• Cold-DM the operators in the verbatim horror-story threads (Naxos captain, RIB operator archetypes are findable in those very forums).

10. Build complexity — justification

Medium. Off-the-shelf: WhatsApp Business API, an LLM for email/slot parsing, a standard web stack, email-ingestion (a forwarding inbox). The custom work is the fuzzy slot-reconciliation engine (resolving differently-named departures to one capacity bucket) and robust multi-template, multilingual parsing that fails safe — a false “all clear” is worse than a false alarm, so the confidence/dedup logic needs care. Gated-API connectors (Viator/GYG) are a fast-follow, not v1. A focused pair ships a credible v1 in ~3–4 months.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Operator forwards their own booking emails / uses their own API access. No scraping of OTA without consent.
Ethical — no harm / dark patterns	✅	Prevents overselling — strictly pro-consumer and pro-operator.
Market exists (evidence above)	✅	39–58% unautomated long tail; verbatim overbooking pain; funded incumbents prove willingness to pay.
1–5 person team can build this	✅	Email ingest + LLM parse + WhatsApp alerts; pair, ~3–4 months.
Launchable with <$50K / ₹40L	✅	API + inference costs only; no inventory, no field ops.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Hair-on-fire at the moment of overbooking — refunds + a rating drop that hits OTA ranking. But felt episodically (a few times/season), not daily, so not a clean 17+.
Demand evidence	15	13/15	Multiple independent signals: hard adoption stats, verbatim horror stories, funded incumbents charging 6–8%. A skeptic nods.
Build feasibility	15	11/15	Off-the-shelf stack but the fuzzy-reconciliation + fail-safe parsing is non-trivial; ~3–4 months.
Distribution clarity	15	12/15	Named, list-able channel (1-star OTA reviewers + operator FB/WhatsApp groups) with a fresh-wound hook. Conversion math still unproven.
Revenue mechanics	15	10/15	Flat pricing is benchmarkable and wallet-friendly, but $29/mo means you need volume; churn risk if an operator’s overbooking is rare enough to forget the value.
Time to first revenue	10	8/10	Free-manifest → paid-tripwire funnel can convert in weeks; DM-the-wounded-operator closes fast.
Defensibility	10	3/10	Thin. Sync is a solved category; the moat is positioning (no-migration, WhatsApp-first, flat price) + parsing data that compounds — copyable in months by an incumbent who chooses to.
Total	100	73/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy (parsing/reconciliation reliability is the product) · content-heavy (distribution is operator-community storytelling in multiple languages).

Key assumptions to validate (3–5)

1. Assumption: Small operators will reliably forward booking emails (or it can be auto-forwarded) so the count stays complete. How to test: 15 operator interviews; have 5 set up forwarding and measure capture completeness over 2 weeks.
2. Assumption: The WhatsApp tripwire fires before the oversell often enough to be worth $29/mo — i.e., email arrives fast enough vs the actual oversell window. How to test: instrument real bookings for 5 pilot operators; measure alarm-lead-time vs oversell events.
3. Assumption: Operators value avoiding occasional overbooking enough to pay monthly (not just at the moment of pain). How to test: pre-sell 20 annual Solo plans off the Loom-DM motion before building the full tripwire.
4. Assumption: LLM parsing hits >98% extraction accuracy across templates/languages with safe failure. How to test: collect 200 real confirmation emails across OTAs/languages, measure precision/recall, especially false “all-clear.”

Risk flags

1. Platform dependency: OTAs could ship faster native sync, change email formats, or restrict forwarding/API. Mitigation: the no-migration positioning + multilingual long-tail focus is exactly where incumbents are slow.
2. Defensibility (low): Channel managers can add a “lite alarm” tier. Speed, niche-language distribution, and flat pricing are the only head start. Must win the segment fast.
3. Market timing / churn: If an operator’s overbooking is rare, perceived value decays between incidents → churn. The daily manifest exists specifically to deliver value between incidents.
4. Value-vs-price: $29/mo demands volume; CAC must stay near-zero via community/DM motion, not paid ads.

14. Structured verdict

Score:                  73/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder who can ship reliable multilingual parsing + a WhatsApp ops layer, paired with someone who lives in operator communities
Time to revenue:        4–8 weeks (pre-sell Solo plans off Loom DMs before full build)
Capital to launch:      $5–10K / ₹4–8L (inference + WhatsApp API + a forwarding inbox)
Top 3 assumptions to validate first:
  1. Email-forward capture stays complete enough to trust the count — 5-operator 2-week capture test
  2. Tripwire lead-time beats the oversell window in real bookings — instrument 5 pilots
  3. Operators pre-pay annual Solo before the tripwire exists — 20-presale target off DM motion
Kill criteria:
  - Abandon if <20% of 50 wounded-operator DMs reply with interest
  - Abandon if alarm-lead-time beats the oversell window in <60% of real overbooking events
  - Abandon if a major channel manager ships a free WhatsApp overbooking alarm before your v1

15. Next step — 1-week validation sprint

• Day 1–2: Scrape 80 recent 1-star OTA/TripAdvisor reviews containing “overbooked/double booked/no room” → extract the operators → record a single generic 60-second Loom showing a mocked WhatsApp tripwire alarm.
• Day 3–4: DM all 80 operators (in their language) with the Loom and a pre-sell: “$199 for a year of SlotSentry Solo, founding price, refundable if it doesn’t catch one for you.” Simultaneously collect 150 real OTA confirmation emails across languages and run an LLM-extraction accuracy test.
• Day 5: Decide. Go if ≥10% of DMs reply with interest AND ≥3 pre-pay AND extraction accuracy ≥95% with safe failure. No-go if replies are tepid or parsing can’t be trusted.

Falsifiable outcome: paid pre-sales count + measured extraction accuracy — not vibes.