This summer I house-sat for friends who went away for three weeks. Their only real instruction had nothing to do with the plants or the cat. It was: “just make it look like someone’s home.” So a few times a week I drove over, opened and closed some curtains, left a different light on, nudged the car forward in the driveway, dropped the mail inside, cracked a window. Ten minutes of theater, then I left.
Standing in their dark living room one evening, flicking a lamp on for an audience of nobody, it struck me how little “presence” actually is. A burglar walking a street at dusk is not checking for people. They are reading signals: a light that changed since yesterday, a shadow crossing a window, a rhythm that says someone lives here and could walk back in at any moment. I was faking those signals by hand. Badly, and only when I remembered to.
The thought that stuck with me was that the single most convincing signal, the one no timer-plug can fake, is a human shape moving behind a curtain. You catch it from the street and your brain fills in the rest without asking permission. So why can’t a window do that on its own?
Curtain Call is a thin rear-projection film you stick to the inside of a window, plus a small projector that throws choreographed silhouettes onto it from behind. From the street it reads as someone crossing the room, pausing at the sink, getting up in the night. Not a looping clip, but scenes with believable timing. And because the interesting part is the performance itself, the scenes are content. You make your own or download them from a shared library.
The house never needed me. It needed the ten minutes of theater I brought with me. Curtain Call is that ten minutes, automated and handed to a timer that actually thinks.
Market & Opportunities
Simulating an occupied home is an old idea with a real market. People already buy timer plugs to cycle their lamps, and FakeTV, a little box that mimics the flicker of a television, sells precisely because a changing glow behind the curtains reads as “someone is up.” The appetite is there. What nobody has done well is the part that actually convinces: a moving human shape.
The clearest buyers are the people who leave homes empty on purpose. Frequent travelers and second-home owners are the obvious core. Renters are an underserved segment right next to them. They cannot bolt a security system to a wall, but they can stick a film to a pane of glass and peel it off when the lease ends, no drilling and no landlord to ask. There is also the quieter case of adult children who want a parent’s house to look occupied and less like a target, without turning it into a surveillance site.
Why now is mostly about cost. Pico projectors are cheap and finally bright enough, rear-projection film is a commodity you can buy by the roll, and the piece that turns a gadget into something people come back to, a library of scenes you can browse and swap, is just software. The hardware has been possible for a while. The content layer is the part that was missing.
Key Features
- Peel-and-stick rear-projection film. Renter-friendly by design: it goes on the inside of the glass, needs no drilling or permission, and comes off cleanly. The illusion lives on a surface you were never going to modify anyway.
- Choreographed silhouette scenes. Not a loop. Scenes carry randomized pauses, entrances, and exits, so a figure crosses, leaves, and comes back later instead of pacing on a visible three-second cycle that gives the whole thing away.
- A scenario library you create or download. This is the heart of it. A dinner being cleared, someone pacing on a phone call, a night owl padding to the kitchen at 2am. Design your own and share it, or grab someone else’s, offered under the same open license as everything here, so the library grows without anyone owning it.
- A rhythm engine. Scenes follow a believable daily arc and sync to local sunset, so nothing “performs” at noon into a bright empty room. The point is to match how a real household actually moves through an evening.
- Multi-window coordination. Different rooms show different figures, so it never looks like one silhouette teleporting around the house. Coverage reads as a family, not a ghost.
- Optional layering. Trigger a smart bulb or a low, muffled murmur of sound alongside a scene, for the nights you want the full performance rather than just the shadow.
Monetization
Three paths, and the first is the honest anchor. Sell the kit, film plus projector, as a one-time purchase, bundled by the number of windows a home wants to cover. This is the version most people just want to buy once and forget.
Second, a scenario marketplace. The community library stays free, because that is the spirit of the thing, but designers can sell polished premium packs (a convincing dinner party, a believable “someone working late”) and the platform takes a cut. Third, an optional subscription for people who want the app to do more: cloud-synced scenes across several homes, seasonal packs, smarter scheduling. Kept deliberately optional, so the base product works fully without it. As a rough sense of scale, a single-window kit sits comfortably in the price range of a mid-tier smart bulb setup, not a security system.
Next Steps: How I’d Build It
The MVP is almost embarrassingly cheap. One window, a phone or a pico projector behind a sheet of rear-projection film, and three or four hand-animated silhouette clips with randomized gaps between them. Then comes the only test that matters. Stand in the street at dusk and see whether it fools you.
The hard part is not the hardware but the craft of convincing. A flat, two-dimensional shadow can look flat, from a sharp angle, under a streetlight, or the moment a loop repeats itself and the eye catches the seam. Getting “random but believable” right, the timing, the scale of the figure against a real room, the way it should never quite repeat, is the actual product. Projection also gets washed out by bright ambient light, so this works best at dusk and after dark, which happens to be exactly when it matters most.
To be clear, this is a deterrent layer, not a lock. It raises the odds that someone picks the obviously darker, quieter house next door instead of yours. It does not stop a determined burglar, and it should never be sold as if it does. Pair it with the boring, real measures. The upside is that nothing here depends on a closed API or a change in the law. It needs a good film, a bright-enough small projector, and a library worth browsing.
Bonus: the most convincing scene is the most boring one
When I tried to picture the scene that would sell the illusion best, it was not anything dramatic. It was a figure wandering to the kitchen and just standing there, lit by the open fridge, doing nothing in particular. Presence, it turns out, is mostly boredom rendered faithfully. One rule, though, sums up the ethics of the thing: keep the performance on your own windows. The goal is to look home, not to project anything at the neighbors.
Closing
Curtain Call is a cheap film, a small projector, and a library of quiet human moments that make an empty house lie convincingly on your behalf. It is yours to build, and you owe me nothing for it.
If you build this, tell me. I would love to stand in a street at dusk and fail to tell whether anyone is home. And if you want a second pair of eyes while you get the illusion right, I am glad to help.
