Cellar Door Cycling — a roving support van for wine-country weekends
Chapters
On this page

Audit of the New World research

Independent reviewer, all six load-bearing prices re-fetched live on 12 Aug 2026. The numbers are sound. All confirmed verbatim on the operators' own sites today: Takaro "from $25p
Cellar Door Cycling — a roving support van for wine-country weekendsresearchingChapter 12 of 76

Independent reviewer, all six load-bearing prices re-fetched live on 12 Aug 2026.

5.1 Every spot-checked number CONFIRMED

The numbers are sound. All confirmed verbatim on the operators' own
sites today: Takaro "from $25pp" delivery and the $35/$25/$35 shuttles · On Yer
Bike's full 15-cell price table and half-day rates · On Yer Bike's $50 per
bicycle
recovery fee · Explore Marlborough $45/$65/$90/$120 plus the wine-pickup
FAQ · Around The Basin $165/$115/$90/$140/$80 and winter non-operation · Wine de
Roads' two quotes and $210/$135 · Trail Journeys $45/$50 per movement with the
15 kg / 9:00am / 11:00–4:00 logistics.

Two drifts: On Yer Bike's site no longer contains the quoted "collect at our
base between 4:00 – 5:00 PM"
(now just "collects it in the afternoon" +
"RETURN BY 5PM") — so the "one afternoon loop" argument is now unsourced; and
Around The Basin's infant equipment renders $74.97, not $75.

The problems are not in the numbers. They are in the tags, the
universal negatives, and one operator counted twice.

5.2 CRITICAL

D-1. "At least four NZ operators run an afternoon wine-pickup vehicle" — it is
three.
The body names On Yer Bike, Explore Marlborough and Wine Tours by Bike.
The fourth in the list is Napa Valley Bike Tours — California. Central Cycle
Trail's "one bottle per 3 pax" is an inclusion, not a collection service. And two
of the three are in Marlborough, so the sample is one region deep.

D-2. "Nobody prices rescue" is contradicted four times.
On Yer Bike prices coming to get you at $50/bicycle; Wine de Roads sells the van
standalone as a dinner driving service from $65; Trail Journeys sells vehicle
relocation at $350/$500
and luggage at $45–50 per movement as standalone
bookable services. Narrow the finding to "nobody sells a subscription or per-day
roving-rescue product"
— which is true, and much less sweeping.

D-3. Explore Marlborough is one business entered twice — and it has been SOLD.
It runs as a live price-verified competitor and, separately, as "a wine-bike
operator currently for sale"
, never connecting them. They are the same business
(homepage <title>: "Explore Marlborough Wine Tours"; both at The Vines Village /
Rapaura Road). The Bayleys listing now shows Sold, so "currently for sale" is
wrong as of today. This corrupts two conclusions: the "realistic shape of a
successful wine-bike business"
callout, and the seasonality chapter's headline
datapoint (the "4 months holiday" disclosure).

D-4. Vine Hopper is a ~75-minute-headway circulator, not a once-a-day loop — and
the chapters disagree.
This one gives one timetable (09:45 out,
16:45 back) and draws the lesson "one vehicle covers three products, rotated by
weekday"
. Vine Hopper's own Northern Route page lists departures at 09:45, 11:10,
12:30, 13:40, 14:50, 16:00
; Southern at 08:45, 10:00, 11:15, 12:30, 13:45.
That is 5–6 loops a day and almost certainly more than one vehicle. The aggregator
research already records the correct Southern timetable and calls it "roughly
75-minute headway".
This is the wrong version — and its incorrect
"one van, three products" lesson has been carried into the International
Synthesis
as a transferable operating model. It must be pulled from there.

D-5. "$25pp means door delivery is affordable to include rather than charge for"
inverts a retail price into a cost.
$25pp is Takaro's minimum retail price,
explicitly "from" and "depends on location and numbers" — the floor for a large
group close to base. It is revenue, and says nothing about the marginal cost of a
two-person delivery 30 km out. Using a competitor's best-case list price as your own
cost basis shows up later as a margin hole.

5.3 MODERATE — mis-tagging is systemic

  • D-6. The Bayleys business-for-sale listing is tagged [VERIFIED], twice over.
    It is a real-estate agent's marketing copy about a third party —
    [SECONDARY] by the scan's own definition. Its "record revenue" refers to the
    2023 season, presented in a 2026 scan with no staleness note.
  • D-7. Aggregator content tagged [VERIFIED] — "[VERIFIED via advcollective
    listing]"
    and Explore Marlborough's "advcollective NZ$68.39" inside a [VERIFIED]
    block. advcollective is an aggregator = [SECONDARY]. NZ$68.39 is an odd enough
    number to be a currency conversion, not an operator price.
  • D-8. [VERIFIED] attached to an absence"No alcohol/liability policy
    published… [VERIFIED] Bikes 'n Wines FAQ — by its silence."
    Silence is
    [NOT FOUND]. "The main cycling operator" is also an unsupported superlative.
  • D-9. This contradicts the regulation chapter on the NZ Adventure Activities
    exemption.
    Here "pure hire likely exempt" sits in a [VERIFIED] row;
    Regulatory and insurance friction says "WorkSafe does not publish an explicit
    exemption list, and this inference is [SECONDARY] — it should not be relied on
    without advice."
    The regulation chapter is right.
  • D-10. [VERIFIED] covering opinions — "the clearest alcohol-risk-driven fleet
    decision in the entire scan"
    , "commercially interesting", "reshape the whole
    economics"
    . The underlying facts are verified; the judgments are not.
  • D-11. Trail Journeys "priced per bag" is inference tagged [VERIFIED]; the
    site says "per movement", and the correct thing is said 12 lines later.
  • D-12. "Own bike $80" is a transport fare, not a support-only price. What it
    buys is the scheduled 9:30am coach Queenstown→Arrowtown and the 4:30pm Gibbston
    pickup — ~25 km each way — plus route access. No rescue, no roving vehicle, no
    wine handling.
    The arithmetic confirms it: $115 − $80 = $35, a plausible
    bike-hire rate, so the $80 is transport priced as transport. Calling it an
    "unbundled support-only benchmark" imports a meaning the operator never sells.
  • D-13. "Wine de Roads is the sole operator promising retrieval in writing" is
    false.
    Takaro's hawkes-bay-trails-info page: "Tākaro Trails provides a pick up
    service for you and your bike if you find you've spent too long in a winery and
    don't fancy the ride home! We'll deliver you safely to your next accommodation
    stop."
    A written promise to retrieve a rider specifically for over-indulging,
    and deliver them to accommodation
    — arguably the closest match to the Cellar Door
    promise in the whole scan, and it was missed.
  • D-14. "Nobody delivers wine to accommodation — that shortcut is what makes it
    free"
    is weaker than presented. Takaro delivers bikes to accommodation and
    riders to their next one; Napier City advertises door-to-door; Bay Cycles
    delivers and re-collects; Sonoma Adventures delivers to "hotels or rentals".
    Several already run accommodation-door movements, so adding wine is not the new
    cost leg implied. Honest version: "nobody advertises it" — a marketing
    observation, not an economic one.
  • D-15. "The Airbnb gap is real and unoccupied" conflates two things.
    Packaging short-lets (nobody does — and the regulation chapter gives the
    good legal reason)
    versus serving guests who stay in them (several already do). Sonoma Adventures'
    own words — "delivery of bikes to local hotels or rentals" — are exactly the
    second
    , cited in one place and ignored in another.
  • D-16. The e-bike premium table mixes units: the Napier row uses
    guided-tour prices ($120/$160) when their e-bike hire is "from $55"; the Takaro
    row uses multi-day against everyone else's day rates. The +33% low end is an
    artefact.
  • D-17. Fleet-value "brackets" use penalty charges as asset values. On Yer
    Bike's "lost e-bike $2,500" is liquidated damages, not capex. Only Takaro's
    NZ$4,500 / NZ$1,000 is a real capex datapoint — and it sits above the
    $3,500 assumption, not bracketing it.
  • D-18. Bike It Now! is tabled as "per-person priced" shuttle, but are also
    said to publish no rate card, with no shuttle price given.

5.4 MINOR

  • "Three continents" appears twice; the scan covers NZ, USA, South Africa,
    Argentina and Chile — four.
  • Getaway Adventures called "34-year-old"; "since 1991" is 35 in 2026.
    Bikes 'n Wines quoted as both "Established 2007" and "20+ years".
  • Around The Basin's product list is incomplete — also sells Half Day (from $105)
    and Full Day Meet Arrowtown (from $100). The latter is the full day without
    the outbound shuttle: a second, cleaner datapoint on what ATB charges for transport.
  • "Three operators refuse e-bikes for related reasons" — only two give reasons
    (alcohol+weight; insurance). Bike 2 Wine simply doesn't stock them.

5.5 Operators this scan should not have missed

  • Sonoma Valley Bike Tours & Rentals (sonomavalleybiketours.com) — live,
    HTTP 200
    , guided and rentals for self-guided, FareHarbor. Entirely absent from
    the California section, the one leaned on hardest.
  • Franschhoek Wine Tram (winetram.co.za, R350pp, five colour lines) — the
    dominant shared-vehicle hop-on-hop-off in the SA winelands, bigger and cheaper
    than Vine Hopper
    , which is crowned "the purest shared-vehicle economics case".
    It is in the aggregator research — a blind spot across chapters, and the
    superlative is unearned.
  • Backroads, Trek Travel, DuVine all sell Napa/Sonoma and Willamette wine cycling
    with support vans and are covered under Region — Global majors. The claim
    "nobody runs a genuinely roving shared van for self-guided customers" is
    drawn from a sample that excludes the operators most likely to falsify it.
  • Willamette is four entries deep, two of which publish nothing — too thin for
    the Oregon conclusions.
  • Wine Country Cyclery — upgrade the evidence. Status is given as "cannot
    confirm". winecountrycyclery.com returns HTTP 404 "Squarespace – Website
    Expired"
    — an expired paid subscription is much stronger evidence of an exit
    than a DNS failure, and belongs in the failures list, not the unknowns.

5.6 The structural criticism — the most important item in this audit

The scan admits its "WebSearch budget was exhausted… operators findable
only by search may be missing."
That is flatly incompatible with the
universal negatives stated alongside it as findings
"nobody prices rescue",
"no operator builds on Airbnb-type inventory", "the only published price found on
three continents"
, "nobody delivers wine to accommodation".

Every one is an absence-of-evidence claim resting on a self-declared non-exhaustive
read of public marketing pages, in a sector the same research documents as
arranged by phone — Takaro's "Call us to arrange", Bay
Cycles, Sonoma Adventures, Bike It Now!'s enquiry-only model.

Services that are never advertised cannot be counted by reading advertisements.
Rewrite these as "not advertised by any operator in this sample" and the research
stands up well. Leave them as "nobody does this" and the business case rests its
central wedge on a search budget that ran out.


Combining both audits. Items 1–5 change conclusions, not just wording.

# Action Source
1 Pull the "one van, three products, rotated by weekday" model from the Synthesis — Vine Hopper runs 5–6 loops/day on ~75-min headway D-4
2 Rewrite the FareHarbor conclusion — the source says a 2026 signup would not pay the 6%; "Australia is in scope" is unsourced C-1
3 Withdraw the €6-vs-€12 load-factor proof — volume discount vs a "from" price; it holds its own counter-example C-2
4 Re-cut the Escapegoat comparison — prices a private guide + Adelaide transfer as "the vehicle" C-3
5 Fix the TrailHopper comparison — like-for-like A$189.95 vs A$189: the same price, not half X-2
6 Requalify every universal negative as "not advertised in this sample" §5.6
7 Narrow "nobody prices rescue" — the research lists four priced support services D-2
8 Merge the two Explore Marlborough entries and mark it Sold D-3
9 Drop the Wine Australia grant from the winery pitch — threshold, closed 2026, paused 2028–30 C-4
10 Correct "at least four NZ wine-pickup operators" to three D-1
11 Re-tag: Bayleys, advcollective, silence-as-verified, opinions-as-verified D-6→D-11
12 Reframe "own bike $80" as a transport fare, not a support-only benchmark D-12
13 Add Takaro's written over-indulgence pickup promise — it falsifies "Wine de Roads is the sole operator" D-13
14 Strip numbers from the Viator and GetYourGuide rows — 2018/2019 third-party; the downside case rests on one §3.4
15 Delete the two self-falsified Gaps in the aggregator research C-5
16 Add the missed operators: Sonoma Valley Bike Tours, Franschhoek Wine Tram; check the majors' self-guided tiers §5.5
17 Verify SA s.47 demerit points at primary source
18 Upgrade Wine Country Cyclery to a confirmed exit (Squarespace expired) §5.5
19 Fix the grossetwines.com https link