I have built over fifty VR projects across Unity, Unreal, and web-based platforms. Puzzle games, horror experiences, racing games, training simulators, branded activations. Each one taught me something about what makes presence click - and what breaks it instantly.
The most important UX principle in VR is comfort, and comfort is not optional. Simulator sickness ends sessions. It does not matter how beautiful your world is if the user feels nauseous after two minutes. I design around the comfort zone first and add visual ambition second. This means locked horizon lines during locomotion, vignetting during fast movement, and never, ever taking control of the camera away from the user.
Locomotion is where most VR projects make their defining tradeoff. Teleportation is safe but breaks immersion. Smooth locomotion feels natural but causes sickness in sensitive users. My default is offering both and letting the user choose in settings. But the real insight is that the best VR experiences minimize locomotion entirely. Design rooms, not worlds. Bring content to the player instead of making the player travel to content.
Spatial UI is the hardest design problem in VR. Flat UI panels floating in 3D space feel wrong. They break the illusion of a physical world. The best spatial interfaces are diegetic - they exist as objects within the world. A wristwatch for health stats. A physical map for navigation. Control panels built into the environment. The UI should feel like part of the world, not a layer on top of it.
Hand interaction design follows a counterintuitive rule: less precision, more forgiveness. In reality, you can pick up a coin with two fingers. In VR, hand tracking is imprecise and latency makes fine motor tasks frustrating. I design grab zones that are 20-30% larger than the visual object. I add magnetic snapping when hands get close to interactable items. I provide audio and haptic feedback the moment a grab is registered. These invisible assists make interactions feel precise even when the tracking is not.
Scale is the most underappreciated tool in VR design. A room that is 10% too large feels empty. 10% too small feels claustrophobic. I prototype environments in graybox first, put on the headset, and adjust scale until the space feels right. This is not something you can judge on a flat monitor. Your body knows when a doorframe is the wrong height. Trust that instinct and iterate in-headset.
Audio is half the immersion budget. Spatial audio - sounds that come from specific positions in 3D space - does more for presence than visual fidelity. A distant footstep behind you triggers a visceral response that no amount of polygon count can match. I invest disproportionately in audio design early in the project. Placeholder visuals with great audio beats polished visuals with flat audio every time.
Performance in VR has zero margin. You need 72-90fps depending on the headset, and every dropped frame is felt physically. I budget my rendering pipeline from the start: 11ms per frame maximum. That means LOD systems, occlusion culling, baked lighting where possible, and aggressive draw call batching. I profile weekly, not at the end. Performance problems are cheaper to fix when they are small.
The meta-lesson from fifty projects is this: VR is not a screen you wear on your face. It is a place you inhabit. Every design decision should serve that illusion. When you forget this and treat VR like a 360-degree monitor, the results feel flat. When you commit to it and design for presence, the results feel like magic.