11. Two structural alternatives

11.1 Asset-light: partner with an existing fleet

Removes: bike depreciation, battery fund, maintenance, theft/damage, fleet material-damage
insurance, storage, the SA/NSW e-bike certification exposure, lithium charging fire risk, and
— materially — you are no longer the supplier of the vehicle when something goes wrong.

Adds: a per-bike-per-day fee, and dependence on a partner's fleet quality on your
busiest weekend.

Market rates — VERIFIED:

EBIT comparison, 66 days:

Crossover rates — the decision rule:

Crossover expressed as utilisation — owning wins when annual per-bike fixed cost
($1,195 ex-GST: depreciation $445 + battery $227 + fleet cover $250 + storage $273) is
beaten by the partner rate net of the $18.18/hire-day variable you avoid:

The recommendation

Do not buy bikes in Year 1. At Year-1 volumes asset-light is better on P&L and
avoids $48,000 of capex — roughly $63,000 of Year-1 cash, on a business whose
peak funding need is $240,000. It also removes the fleet cover line from the insurance
stack in the year the insurance is hardest to place.

Then the decision is entirely a negotiation. Go to a bike-hire partner with a
seasonal block booking and find out whether they will do $40/bike/day. If they will,
never own — you would need 99.5% utilisation to beat it. If the best available is
$70+, own from Year 2, once utilisation clears 40%.

This is a phone call, not a model. Make it before spending $32,000.

11.2 Cellar-door referral revenue — the honest answer is "not yet"

The commission model does not work at this scale, and the reason is arithmetic.

Model it generously: average rider buys $120 of wine; the van collects $80 of it; a 10%
commission on collected purchases = $8/rider. At 792 riders (Year 3) that is
$6,336/year — about 7% of contribution. Real, but not transformative.

The kill is distribution, not rate. 792 riders across 12 cellar doors is
66 riders per cellar door per year — roughly 1.3 riders per weekend. No winery will
build a till process, train staff on a rider code, and reconcile a commission statement
for 1.3 customers a weekend. Australian cellar doors run thin margins and many pay tour
operators nothing at all as a matter of policy.

What would have to be true for the commission line to work:

1. ~5,000 riders/year — enough that each cellar door sees 8–10 riders per weekend and
   the process pays for itself. That is 4–6× the Year 3 volume, i.e. multi-van, multi-region.
2. Clean attribution — a rider code at the till, which is friction at exactly the
   moment you want none.
3. A winery that believes the counterfactual — that these sales would not have happened
   anyway. Riders arriving on the trail look like riders who would have come regardless.

The alternative that does work at small scale — a flat trail-partner fee. $500–$1,000
per cellar door per year for inclusion on the route card, wine-collection service, and
signage. 8 cellar doors × $750 = $6,000/yr — the same money, but sold as a marketing
line rather than a rebate on a sale they think they already had. Far easier to sign, and
it does not need attribution.

Model both at $0 in the base case. ESTIMATED, no basis in any quote. This is the
cheapest thing in the whole plan to test: call five cellar doors. It costs an afternoon.

The far better second revenue line is already sitting in §4.3 of the bike-fleet
chapter: the damage waiver. $12/bike/day at 50% attach on 792 riders = $4,752 of
Year-3 revenue against ~$1,200 of extra retained loss. It is comparable to the
referral line, it requires no third party to agree to anything, it is standard practice
across every Australian bike-hire operator reviewed, and it is collected at the same
moment as the booking.