Fifteen years of work already built the hard part — the name, the network, the access, and an archive nobody has published. What has never existed is a company underneath it.

Every film this company makes, it owns. That is the whole thesis. A film delivered to a client in spring goes to festival tours that autumn, licenses to a broadcaster the year after, runs on ad-supported platforms indefinitely, and sells into an AI training market now paying $60 to $240 an hour for footage. Same shoot, same crew, same budget — four times the life.

The adventure space ran on corporate media for thirty years, and those structures are in decline. Every institution that held a piece of it has contracted, been sold or been broken up in the last four years, while participation hit a record 183 million Americans. What is filling the gap is not another magazine or another network. It is creators running their own companies and owning their own output, and almost nobody has done that in adventure at any real scale.

• Months 1–2. Clear the ledger. Entity, contracts, insurance. Crew in place. Gear and kit bought.
• Months 2–4. The archive comes online and starts publishing. Inbound brand pipeline reopens against a current reel.
• Months 4–6. First departure runs, partner-funded. First owned film delivered.
• Month 6 onward. Operating profit. Each film funds the next departure.
• Year 2. Cadence doubles. Library compounds. Licensing becomes a standing revenue line rather than an event.

Expeditions are productions, not tours. Partners fund most of what they cost to make. The company keeps the films.

Structured as a committed facility rather than a lump sum. You commit; the company draws only what it needs against a forecast. Only drawn capital is repayable. If less is needed, less is owed.

Repayment is a share of free cash flow, not a fixed schedule — so it never competes with the company's ability to grow into the thing that repays you.

Patronage, principal-only, a fixed return, revenue share to a cap, or equity. All are on the table. Terms are settled in conversation.

This is a long hold. Principal comes back in years, not months. Anything needing liquidity inside three years does not belong here.

Output scales with capacity. At this number the slate runs on one director and a small crew, which caps how much gets made in a year. Bringing other athletes and filmmakers under the company is what the numbers above this one buy, and it is the difference between four films a year and twelve.

Licensing compounds, it does not arrive early. The film tail and the AI training-data line are both real and both lumpy in the first two years. Nothing in the plan depends on either landing on schedule.

A full three-year model sits behind every figure on this page. Ten tabs, roughly 700 formulas, and a source citation for every benchmark used. It is built to be argued with rather than presented.

What is in it. A cost build covering people, production, gear, insurance, legal and accounting, with employer loading applied properly rather than assumed away. A revenue build by channel, each line driven from its own inputs rather than a single growth rate. A profit and loss and cash statement across three scenarios. A drawdown and repayment schedule out to year seven. And an odds tab covering the base rates for what fraction of people who attempt this actually reach each level.

What it currently says. Three-year cost of $1.68m against $2.27m of base-case revenue, so the business is net cash positive across the period. The cash trough is $319,000 at the end of year one, which is where the $500,000 comes from rather than it being a round number. Peak capital actually drawn is $207,000 against a $500,000 commitment, and the drawn balance is repaid by year five with no speaking income at all, year three with it.

What it does not do. It does not invent figures where none exist. Outdoor-specific platform rates, documentary licence fees at this length and festival tour fees have no reliable public data, and the model says so in each case rather than filling the gap. Every assumption is an input you can change and watch the answer move.

Not published here. Ask and it is yours, along with an hour to take it apart.

The underlying financial model runs three years in full detail, with drawdown and repayment schedules out to year seven. Years four and five here extend the same channel build. Available on request.

Red Bull Media House was founded in 2007 — roughly 1,250 events a year across 100+ disciplines, some 700 contracted athletes, distribution into 160+ countries, a library past 600 titles. Adventure is the larger audience and has no equivalent.

There is no adventure-athlete-owned production company operating at scale in Robinson’s lane — none with disclosed financials, institutional backing or a valuation. The model is already proven next door: Dude Perfect raised over $100m against roughly $50m of revenue; Yes Theory built a ~10-million audience and an 18-person company across ten countries on creator-owned adventure content; and Teton Gravity Research turned specialist action-sports media into partnership budgets of $15,000 to $3 million-plus. None of them own the highline.

The eight channels in the model are what the company earns on the plan. These are the avenues that open because it owns the work — none of them reachable as talent-for-hire. Named, not yet modelled; each one is an argument for keeping the copyright.

Named as plausible, not yet priced — the same discipline as the rest of the model. The ones that fit best are where the footage compounds and the rights stay home: format licensing, the series-and-sales-agent pathway, and destination commissions.

At the top of the scale, the company stops renting and builds its own home. The Stokehouse is a permanent base — the thing every number below this one is quietly building toward. It turns a founder-led operation into an institution with an address.

Everything under $5m rents the room. This one buys the house.

These are productions, not tours. Costs assume existing gear, in-house post and network crew rather than hired-in production. Partners fund them. The company keeps the films.

• Platform revenue — built from views and RPM by surface, not a single guessed figure. YouTube long-form $3–7 RPM net of the platform cut; Shorts at 1.5–3% of that; TikTok Creator Rewards $0.40–1.00 per 1,000 qualified views. Instagram is modelled at zero because no per-view payout exists.
• Brand pricing — priced buy-side on views at $15–32 CPM with a floor, per agency rate cards. Not follower-tier cards, which run well above what is actually transacted; marketplace data across 21,000 collaborations shows a $193 median.
• Expedition partners — brand-funded expedition work of this length is commissioned at $54k–250k. The figures here sit at or below the bottom of that band.
• Licensing — no current public rate card exists for a 20–30 minute independent film; the only published one is from 2011. The conservative column is the one to argue from.
• AI training data — Reuters (2024) reported $100–300 an hour for longer films; Bloomberg (2025) reported $1–4 a minute for unpublished creator footage. Lumpy and deal-timed. Nothing here depends on it.
• Speaking — adventure-athlete keynotes run $15–40k, and the keynote-with-live-highline-performance sits above that as a product no competitor can match. The model runs the line conservatively, on realized fees below those rates. Bureau commission 25–30%.
• Where the data does not exist — outdoor-specific YouTube RPM, documentary licence fees at this length, and festival tour licence fees are all unpublished. The model says so rather than inventing a number.

Every benchmark above carries a URL in the full model.