← Back to workStudio concept · 2026

MONOLITH

A self-set brief for a premium members-only gym selling keys before the fit-out is finished — designed, built and shipped end-to-end, the same way a client project runs.

Year
2026
Role
Concept · Design + Build
  • 3D
  • WebGL
  • Landing page

Studio concept

MONOLITH is a fictional Birmingham gym — there's no paying client behind it. The brief was set internally to answer one question: can real-time 3D replace photography for a space that doesn't exist yet? Pre-launch fitness sites almost all fall back to stock imagery and 'artist's impression' JPEGs; this explores the alternative. The deltas below are what would change on day one of a real engagement.

The brief

MONOLITH is a concept brief for a members-only bodybuilding gym opening in a Digbeth railway arch — a pre-launch site that has to sell £89/month keys before the fit-out exists, using a real-time 3D fly-through of the space instead of photography that can't be taken yet.

Live site
monolithgym.co.uk
Open ↗
View the full siteOpens in a new tab
Walk them through the gym the way I would — then show them the price.
From the brief
Design decisions

The four specific calls that defined this build.

The render is the pitch

A pre-launch gym has no photography, and stock iron looks like every franchise in the country. The whole site is one stylised 3D scene built in code — the space itself becomes the brand asset, and it weighs a few kilobytes instead of a photoshoot.

One scroll, one shot

Every gym sells memberships the same way: someone walks you across the floor, then names the price. The scroll is that walk — a single unbroken camera move from aerial shot through free weights, machines, and the recovery suite, landing on the membership grid.

Pricing at the end of the walk

The tour hands off straight into three named prices — £15, £59, £89 — with no enquiry form and no 'from' asterisks. If the walk-through did its job, the price is a confirmation, not a negotiation.

Brand by subtraction

Premium here means what's missing: no classes, no cardio theatre, no gamified app. Near-black, one ember accent, one condensed typeface. The house-rules section does the positioning that a 'why us' page normally fumbles.

Trade-offs

What was ruled out, and why.

No photography.

The entire site is one procedural 3D scene. When the real fit-out exists, film replaces the render — but pre-launch, a stylised model of the actual floor plan is more honest than stock photos of someone else's gym.

No classes, no timetable, no app.

The premium positioning is subtraction: a keyholder gym sells access, not programming. Cutting the timetable, the trainer bios, and the booking flow isn't a missing feature — it's the pitch.

WebGL as the hero, HTML as the floor.

The fly-through is the selling moment, but reduced-motion and no-WebGL visitors get the full content as stacked sections. The 3D is a layer on top of a complete page, not a gate in front of one.

The stack

Every tool chosen on purpose, with a reason that beat the obvious alternative.

Framework
Next.js App Router

The 3D scene ships as a client-only chunk behind a dynamic import; everything else is static HTML.

3D
React Three Fiber + drei

The gym is procedural geometry built in code — no GLB downloads, no textures. Instanced meshes keep the whole building under 80 draw calls.

Scroll
Lenis + Framer Motion

One shared scroll-progress value drives the camera spline, the zone lighting, and the copy crossfades, so nothing can drift out of sync.

Hosting
Vercel

Static-first deploy; the only heavy asset is a ~150KB JS chunk that loads behind a poster.

Analytics
Plausible

Cookie-free. The one funnel that matters: tour completion → 'request a key'.

Fallbacks
Reduced-motion + no-WebGL static build

Every word of the tour exists as plain stacked HTML when the camera can't fly. Content parity, not a 'please rotate your device' apology.

What success looks like

The numbers a real engagement would be scored on.

1 shot
Hero to membership, no cuts
<80
Draw calls in the 3D scene
0
Stock photos — the render is the site
At launch

What would change on day one of a real engagement.

  • Swap the stylised scene for the architect's actual floor plan, or a photoreal Blender fly-through baked to a scrubbed frame sequence
  • GoCardless direct debit behind 'Request a key' (currently a mocked mailto)
  • Access-control integration (Kisi or Salto) so the 24/7 keyholder flow is real, not copy
  • Real photography of the finished fit-out layered into the membership section at launch
  • Plausible funnel: tour completion → key request → direct-debit mandate

Like this for your business?

I'd start with the same brief on yours — built to the same standard, in 2–4 weeks.