What's the best experience you've done this year?

Who's going to create the best one you do next year?

People are moving from things to experiences, and corporate spend is following.

Ask the first question and actually wait. Everyone has an answer, and nobody's answer is a thing.
Scroll

Pre-seed · £250k SEIS

Three Doors

Star in your own spy thriller. Walk out knowing the real you.

An AI-first company, building the most human thing there is.

Tom Ball: Tech agency founder  ·  Builder of physical themed spaces: DeskLodge, £4m t/o, profitable, 12 years  ·  Prototype hardware working today

What a player says the next morning: "A flower shop recruited us as spies, and by the end it knew us better than we knew ourselves." The AI-first line pays off on slide 5, the economics on slide 8, the founder story on slide 9.

03The big idea

An emotional high: starring in your own Hollywood movie.

  • Tricks. Twists. Live actors. Real dilemmas, in private, with your own people, and no audience.
  • The world's best rooms already deliver this feeling. They have never been seen in the UK.
  • It lives where three crafts overlap, and it takes every ingredient at once, tested and re-tested underneath.
Spectacle Story Agency A real world Sound, light& senses Charactersyou believe An endingthat lands Surprises &physical twists Witnessedmoments Stakes thatare yours The Real You
Each circle alone already exists: theme parks own spectacle, immersive theatre owns story, games own agency. Nobody holds the centre. That's the product. Press right for the feeling in 90 seconds and the honest scorecard against the world's best.

04The opportunity

Billion-pound theme parks at one end. Hobbyist escape rooms at the other. Very few between.

1,500+ UK escape rooms £25 to £35 a head £1bn+ theme parks £100m+ per ride This is where the £100m businesses are being built £80 to £200 a head · especially outside London
  • The middle stays empty because it's genuinely hard to create. That's our moat, not our problem: slide 9.
    • Escaparium: the world's #1 room at 5x their own standard prices
    • The Escape Game: 50+ venues, $61m revenue
    • Meow Wolf: five permanent US cities, $200m+ raised
Meow Wolf

Meow Wolf · click to cycle

The barbell is the market: two crowded ends, an empty £80 to £200 band that Belbin-style workshops (£175 to £270, classroom) only pretend to occupy. Demand doubled to £1.4bn in four years. The window is dated: production-grade AI narration 2025, Netflix House Nov 2025. Nearest neighbours exist (Aftermath at £39, SPYSCAPE, Sandstone) and none holds all four dimensions; see the panel.

05The thesis

An AI-first company in a non-AI world.

Hollywood in front of the curtain. Server racks behind it. We sell the most human thing there is, and build it with the most leveraged company we can construct.

  • More immersive · Endlessly repeatable · A fraction of the cost.
  • We believe we can use AI better than any competitor, and the customer never sees it.
A warm film set with actors in front of a crimson curtain; racks of servers glow behind it
The customer never meets the AI. That's the point. Behind the curtain it's the narrator, the stage manager, the analyst, the test lab and half the engineering team. This is the cover's promise, cashed. (Image is a concept visual; we can reshoot with our own set later.)

06FRINGE, the behavioural read

Every player leaves with FRINGE: a personality test scored on what you did, not what boxes you ticked.

  • The incumbents (Belbin, MBTI, DiSC) ask you questions. FRINGE observes you under real pressure: what you said, what you chose, when you moved, how your heart behaved.
  • Who people claim to be on a Tuesday vs who they were on Saturday night.
  • It's why corporate pays up to £200pp for the same 90 minutes.
The player kit on the walnut desk: wired headset, phone and heart-rate armband
The visual is the capture kit every player wears: wired headset (voice), phone (position and orchestration), heart-rate armband. The workflow (experience, debrief, report, revisit) is the moat, the way the Bloomberg terminal habit is the moat. Next slide: the data it captures, live. The folio report now closes the deck.

07Already being built

The prototype is live. Every session is a complete game record.

The Player Tracker: live position in the room, four signal traces over time, and per-player utterances

Our live-ops console, today: where each player stands, their heart rate, voice, and every word transcribed. Four streams, one timeline, every five seconds.

The transcripts wall internal tool

The first test space, being fitted out near Reading right now, and the transcripts wall watching it.

Best-in-the-world is built by testing: overinvest before launch, save 10 to 100x after. We studied the best in the world (121 rooms played, the global Top 100 torn down, 363 podcasts indexed), then built the machine that tests like them.

This is the working prototype (demo data shown; the real pipeline streams from our handsets today: mic to Whisper, heart rate from the armband, position from beacons). Every effect gets proven in plywood before a penny of venue capex; every run becomes a reviewable artefact. No other operator on earth has this view of a session.

08Exceptional economics

An exceptional experience unlocks exceptional economics.

The multipliers compound: price per head * throughput * occupancy, then * sites. Same fundamental format, but a different episode, a different story.

Both columns are saved scenarios in our live model. Press right and move the sliders yourself.

Traditional escape roomThree Doors
New team every90 min12 min
Average teams per week~60*~200
Blended price per head£25£75
Square foot per room500 sq ft10,000 sq ft
Occupancy50%60%, two markets
Venue revenue per year£270k*£3.15m
Operating marginnear zero39%

*Assuming three escape rooms at one location.

Read the table bottom-up: the escape room loses money at 50% occupancy; the same square footage runs at 39% margin when the multipliers stack. Corporate is the price dial AND the occupancy dial at once; that's the dual market. The pathway to £100m is sites x this machine. [PH: pathway figures once the financial plan locks.]

09Why us · Tom Ball, founder

This is hard to create. That's the moat. And I have a history of delivering exactly this.

✓Live hospitality
✓Property
✓Hardware
✓Software
✓Narrative
+Behavioural science
✓Marketing
Live hospitality: DeskLodge, 12 years · corporate parties for 1,000 in my 20s Property: 48,000 sq ft of themed space found, fitted out and filled, profitably Hardware: built the player firmware myself; streaming AI voice today Software: tech agency founder; shipped products people actually used Narrative: 121 rooms played worldwide, with a notebook Behavioural science: FRINGE authored; advisor + validation funded this round Marketing: brand, cast and channel plan built and locked pre-launch
  • Hobbyists have the love but not the capital. Corporates have the capital but not the love. Holding this whole chain is the barrier, and it's why the middle of the market stays empty.
  • Until Claude Opus 4.8 (2026), the economics of this many moving parts did not work. Now one AI-first founder ships what took a department. First mover on that unlock, in the least-AI category there is.
Bold version, out loud: I deliver. Parties for 1,000 in my 20s. A tech agency in my 30s. DeskLodge to £4m and profitable in my 40s, and it's still running. Three Doors is the same muscles pointed at one obsession, and the AI unlock arrived exactly when I was ready to use it. Cockroach, not unicorn.

10The team

A tight, AI-powered core, surrounded by experienced experts.

The core is the founder plus two full-time founding hires: a Game Master / Physical Operator who owns the build-and-test cycle, and an Operations Manager who runs the machine. AI leverage is why three people can do what fifteen did five years ago.

Around them, senior specialists on tightly-scoped paid projects, each engaged with a view to a full-time seat as we scale. The two credibility-critical seats (behavioural science [PH: name], construction) are named before the seed closes.

Seed unlocks execution, not recruitment.

The core founder + 2 full-time hires AI-powered Behavioural-science advisor Experience designer L&D channel opener Construction lead Marketing strategy Social content Creative pair AI-leveraged developer
The answer to "is this a one-man band?": a deliberately tight core (three people, EMI options, AI-leveraged) with eight experienced experts around it on scoped paid projects, and the two credibility-critical orbits named before the seed round closes. Same thesis as slide 5, applied to headcount.

11The Disney of Escape Rooms

This is episode one. We're building a universe.

Each experience stands alone in a single location. You do your local one, then travel for the next. Future locations are future episodes, then whole new series.

The Directorate · Episode 1
🔒 The next Episode · earned
🔒 The next series · Mission to Mars?
FRINGE · three-door mechanic · tech stack: episode-agnostic
The map is deliberately unmarked: a network, not a promise. The crimson dot is venue one. The universe point, out loud: nobody retells "we visited a chain". They retell "we've done three episodes and Mars opens in Bristol". Studios earn a slate one hit at a time; Disneyland opened with one park.

12The bet

To believe this is a £100m company, you only need to believe four things.

1 · People will love it

The recipe is proven by the world's best (slide 3); the testing engine (slide 7) exists to iterate until the gasp is reliable.

2 · They'll pay £50 to £150 a head

The band is empty (slide 4), the world's best already charge more, and corporates pay £175+ today for a classroom.

3 · The machine fills

Two markets want different hours of the same week; 60% blended occupancy is the category norm, not a stretch (slide 8).

4 · The margin survives contact

AI carries the cost structure (slide 5); 39% modelled at venue one, and the sliders are yours to move (slide 8).

That's the whole bet. Everything else (FRINGE validation, new episodes, licensing, the measurement layer) is upside on top, not load-bearing.

The recap that doubles as risk framing: each belief points back at the slide that evidenced it. If they nod four times, the ask on the next slide is already agreed.

13The ask

£250k SEIS: prove the format before we're on-site.

Phase 1 · this round

Build + test in a plain warehouse until it's exceptional.

  • Tech proven
  • Format evolved
  • Market proven
Phase 2 · seed

Build the venue against a proven, tested model with proven demand.

SEIS relief · a founder all-in · the cheque removes the two biggest risks before the expensive capital goes in.

The biggest thing this round buys is not a venue. It's the certainty that the venue, when built, works.

Who are you really?

Three Doors

Tom Ball · tom@threedoors.co.uk · Three Doors Experience Ltd · CRN 16748951

The deck now closes on the folio: the thing every player leaves holding, with the question as real typography over it. Everyone who walks out gets an answer. I'd like to build it with you behind the glass.