Distribution — the channel findings that bite, and the gaps stated plainly

From a parallel distribution scan of 13 platforms; the fuller treatment is in the
Distribution And Take Rates — channels, OTA fees, payout terms chapter. Its
headline caveat: none of the 13 platforms publishes a headline commission rate on
its own site, so most percentages there are third-party and dated.

1. You cannot run a minimum-numbers policy and stay on GetYourGuide

Verified in GetYourGuide's own supplier help centre (article 29374073400989,
Performance Quality Standards):

Supplier Cancellation Rate (non-force majeure): partners must maintain a cancellation
rate of 1.0% or lower. No-Show Rate: 0.2% or lower.

90-day rolling. "Failure to consistently meet these standards may result in the
temporary removal of activities from the platform."

This is the sharpest operational constraint found in the entire scan. The model
assumes a ~$755/day van needing ~6 riders to break even, with the implied
fallback of cancelling thin days. Cancel more than one booking in a hundred and you
are delisted. Any "we run if we get six" clause is incompatible with OTA
distribution — and yet quorum dispatch (Trans Cascadia, Hike n Bike, Hoodoo, Slow
Days) is the industry norm.

So there is a real either/or to decide deliberately: run for one couple and skip
quorum, or use quorum and stay off the major OTAs, selling direct and through Airbnb
Experiences. Buller Adventures' per-vehicle pricing is the third way — the small
group pays for the empty seats and nothing ever gets cancelled.

2. Airbnb Experiences is the most aligned channel available, and it is 20%

The only on-the-record own-site commission rate in the entire distribution scan:
Airbnb help article 1857 — "For experience
reservations, we typically charge a 20% service fee." Corroborated at 20% by
Checkfront (2020) and Tourpreneur (2020).
Caveat: the sentence sits in the host-fee section but does not literally say "host",
so the payer is verified-with-caveat.

The customer is already sitting in an Airbnb in the region, and Airbnb surfaces
experiences to guests who have booked stays nearby. No other channel matches intent,
geography and timing that precisely — and it sidesteps the short-let-host referral
problem (fragmented, low-loyalty) by going through the platform the hosts already use.
Format is short/single-day, which fits a weekend day ride exactly and rules out
multi-day touring.

3. Take rates, such as they are

What this does to the model. The draft assumes $40/rider CAC. On a
$189 product, a 20% take is $38 and a 30% take is $57. So $40 is roughly right
at the optimistic end and understated at the realistic end — and it is a commission
off the top, not a marketing spend you can throttle. At 30%, contribution per rider
drops from $124 to about $107 and break-even rises accordingly, from ~6.1 riders
per operating day to about 7.1.

4. Other verified operational details worth pricing

• GetYourGuide pays on TRAVEL DATE, not booking date — invoices include "bookings
  with travel dates in the previous month". Monthly: invoice 1st business day, payment
  5th, bank transfer up to the 12th. Bi-weekly payouts cost +2% commission. For a
  weekend product with short lead times this is a mild drag, not the serious one it
  would be for multi-day trips booked six months out.
• TourRadar pays on booking due dates, not travel date — "no drip-feeding, no
  minimum thresholds, no rolling or fixed reserves". Structurally the best cash flow
  of the set, and it has a dedicated multi-day cycling category (/f/bicycle).
  Relevant only if a multi-day tier ever exists.
• GetYourGuide listing eligibility risk: "We cannot accept resellers or
  aggregators, online travel agencies, destination management companies, or unregistered
  private guides." A product built on partner-supplied bikes and third-party
  accommodation needs to present as an operator — which it is, since it runs the van
  and owns the route — but confirm before building a channel strategy on GYG.
• GYG commission varies by pricing category within a product (adult vs child rates),
  not one number per supplier.
• Listing is free on Viator, GetYourGuide and TourRadar — no setup or monthly fees.
• Connectivity is not mandatory on GYG: 100+ partners (Bókun, Ventrata, FareHarbor,
  BookingKit, Palisis), direct API, or manual.

5. Not marketplaces — don't model them as channels

• komoot — no tour resale, so no commission. Multi-day
  is explicitly a planning tool with no booking. The B2B offer is advertising: Partner
  Profile (free), Community Expert, Sponsored Collections. A media buy, not a take
  rate. Pricing not published.
• Bikemap — no bookings at all. Consumer
  freemium: €9/month, €39/year, €149.99 lifetime.
• Outdooractive — SaaS content and marketing
  tech for destinations and tourism boards, not a marketplace.
• Tourlane — a tailor-made agency with no supplier surface; you would be a contracted
  supplier on negotiated net rates, not a listing.
• Evaneos — "at least 85% of the trip price benefits local communities". Do not
  convert this to a 15% commission — "local actors" covers hotels, guides, drivers and
  transport, not agency retained margin. Selective: "only 1 in 10 is selected", audited
  code of conduct.
• Much Better Adventures — host knowledge base is login-gated; rate unobtainable.
  One email to their host team would close it.

Gaps — stated plainly

1. Viator's commission is unverified — every supplier-side URL is a hard 403 from
   this environment. GetYourGuide's is unpublished by design. Klook, Civitatis,
   Musement, Evaneos, Much Better Adventures, Tourlane, Komoot, Bikemap and
   Outdooractive were not reached. Industry press (Skift, Arival, PhocusWire) was
   never searched — the budget ran out before discovery could run. Flagged as the
   highest-value thing to redo with a fresh search budget.
2. Nobody publishes vehicle capacity or customers-per-van. Across ~20
   luggage-transfer and shuttle operators — and, separately, ~45 MTB shuttle and event
   operators — not one states how many bags or customers share a vehicle. The load
   factor, the single number this model turns on, is an industry-wide trade secret.
   The only public proxies found are the Brigantes 1-bag/2+-bag split, the Escapegoat
   2-to-10-guest curve, and Mountain Bike Rotorua's 5/9/24-seat charter table.
   Only four disclosures of capacity exist across ~40 operators: Escapegoat (8/11-seat
   vans), Tastes of the Hunter (11/20/24-seat buses), Share Shuttle (5-seat Prado /
   13-seat minibus), Barossa Bike Hire (80+ bike fleet).
   ⚠️ FALSIFIED BY THIS FILE'S OWN CONTENTS — there are at least seven, not four.
   Add Michigan Coast to Coast ("room for 55 riders"), MTB Rotorua (5/9/24),
   Adrenalin Uplift (16-seat bus) and Buller Adventures ("up to 5 passengers").
   Note also that the Van Economics chapter states "exactly four" over a
   completely different set of operators — the number was copied across two
   disjoint datasets, and both counts are wrong. See C-5 in the Audit of the
   aggregator and shared-support research chapter and X-3 in the 7. Cross-file
   consistency audit chapter.
3. No response-time SLA exists anywhere in the world, across every sector examined —
   Camino, UK trail baggage, rail-trail, roadside-assistance subscriptions, MTB uplift,
   event SAG.
   ⚠️ FALSIFIED BY THIS FILE'S OWN CONTENTS. This file records ADFC Pannenhilfe:
   mobile repair up to €20 "within a maximum of 120 minutes" — a published
   response-time maximum, in "roadside-assistance subscriptions", one of the very
   sectors this gap names. Add the hard delivery windows also recorded here:
   Correos "before 14:30", Trail Journeys "delivered 11:00–16:00",
   Travel-Lite/Highland "collected by 9:00am". The 3. How support is actually
   delivered chapter also names Discover France's "back on the road within an
   hour".
   The closest anti-promises remain ETA's "24-hour recovery service" and one
   anecdotal RACV testimonial ("they were here in 20 min").
   Publishing one would make this the first operator anywhere to do so — genuine
   differentiation and genuine liability. That claim is false. The accurate
   version: no operator publishes a response-time commitment for on-route rider
   rescue in a wine region. The liability half of the warning stands; the
   "first anywhere" half does not. See C-5 in the Audit of the aggregator and
   shared-support research chapter.
4. Correos does not publish a per-stage price on its own site; the €4–7 range is
   third-party, plus Correos' own blog examples.
5. Laka's contribution cap and fee percentage are not published — including in their
   own help-centre articles literally titled "What is the Laka Fee?" and "How do you
   calculate my monthly share?".
6. US rail-trail shuttles (GAP, C&O, Katy) quote privately — no rate cards exist.
   Coast to Coast Packhorse could not be located at any working URL.
7. World Expeditions / UTracks agent commission exists on a page geo-fenced to
   North America — retrievable from a US or Canadian IP. The one live thread for
   an Australian trade-commission number.
8. No hard evidence either way on wineries subsidising shuttles. Franschhoek Wine
   Tram discloses nothing about estate participation terms.
9. Whether Airbnb's 20% is host-side is verified-with-caveat, not settled.
10. Still open in the MTB/uplift set: Whistler season pass · Thredbo MTB day pass ·
    Antur Stiniog standard rates · BikePark Wales weekend/annual · Maydena bus seat
    count · organiser cost per day for neutral service · riders-per-van at every Derby
    operator and Yuba Expeditions.