The Hardest Part Is Picking One Thing
Most failed microhorror films fail for the same reason. They try to do too much. Two monsters. Three plot twists. A backstory that needs explaining. You've got under three minutes. Pick one thing and do it completely.
Your concept should be expressible in a sentence. Not a paragraph. Not a pitch. A sentence. "A woman realizes the reflection in her mirror is on a slight delay." That's a microhorror film. "A family moves into a haunted house and discovers the previous owners' dark secret which connects to an ancient curse" is not. That's a feature idea and a bad one at that.
The constraint is a gift. Use it.
Location
One location. Ideally one you can access for free and control completely. Your bathroom. A stairwell. A backyard at night. A specific hallway in your apartment building.
Location is character in microhorror. The space itself should feel wrong, or normal enough that when something wrong enters it the contrast lands. You don't have time to establish multiple spaces. Pick the one that serves the concept and learn every angle of it before you shoot.
Think about light. Where does it come from? Where doesn't it? A bathroom with the light off except for a phone screen is already doing half your work for you. A kitchen at 3am with the single overhead bulb on is a different film than the same kitchen in daylight. The space tells you what the film wants to be if you listen to it.
Sound Design Is Where Microhorror Lives
This is the one area where most first-time microhorror filmmakers underinvest, and it's the one area that matters most. You're not going to have jump scare orchestral stings. You can't afford them and they're not what makes short horror work anyway.
What works: silence that's slightly too long. Ambient noise that's slightly wrong. A sound you recognize in a context where you shouldn't be hearing it. The scrape of something that shouldn't be moving. Breathing that doesn't match what you're seeing.
Spend as much time on post-production sound as you do on shooting. Then spend more. The picture can be rough. The sound cannot.
Practical tip: record room tone for every location for at least five minutes. Silence in microhorror is built from room tone. Without it, your "silence" has all the wrong sounds in it and the spell breaks.
The Twist vs. The Gut Punch
Microhorror doesn't need a twist. It needs an ending that lands.
There's a difference. A twist recontextualizes what came before. A gut punch confirms what you were afraid was happening. Both can work. The gut punch is more reliable because it doesn't require the viewer to do mental backflips in the last thirty seconds of a film that's already ending.
Ask yourself: what is the viewer afraid is true? Your ending should either confirm it in the worst possible way, or subvert it in a way that's even more frightening than confirmation would have been. There's no third option in a good microhorror film.
Don't explain the ending. Don't have a character say what's happening. Show the last image, cut to black, and get out. Trust the viewer. Trust the concept.
The Rule of Threes for Short Horror
Establish the normal state of things. Introduce the threat. Pay it off. That's the whole structure of a microhorror film and you can do it in ninety seconds if you know what you're doing.
The establishment doesn't need to be long. Five seconds of a person alone in an apartment tells you what you need to know. Ten seconds of normal behavior before something shifts. The threat introduction should feel like a door clicking open rather than a jump scare. And the payoff should arrive before the viewer has had time to fully prepare for it.
Timing is everything. If you run the normal state for too long, the viewer gets bored. Too short and there's nothing to violate. Watch "Lights Out" three times and clock exactly how long Sandberg spends on each beat. He's near-perfect.
Keep It Under Three Minutes
You said under five. Make it under three. Under two if you can. The shorter the better, all else equal. Every second you add is a second you've decided you need. Demand a reason from yourself for every additional second.
Cut your first cut by thirty percent. Then ask if you can cut it again.
The Distribution Question
Once you've made it, where does it go? YouTube is obvious but the noise floor is enormous. Vimeo has better production quality audiences. Film festivals for short horror are more accessible than you think and winning one of them, even a small one, changes what doors open for you.
Platforms like FinTV that are built specifically for short-form horror are worth pursuing. The audience on a dedicated horror streaming platform watches differently than a YouTube browser. They're there for the experience, not stumbling past it. That matters for how your film lands.
Make the film first. Make it right. Then worry about where it goes.