Two weeks before Halloween, something interesting happens to every neighborhood: it reveals itself. Some blocks go full haunted mansion — twelve-foot skeletons, fog machines, lawns turned into graveyards. Others do absolutely nothing. No pumpkins. No cobwebs. Just houses sitting there like it's a Tuesday in February.
Parents know this. Anyone who has ever planned a trick-or-treat route knows it. You walk the neighborhood on October 30th, you figure out where the candy is, and by November 1st that knowledge evaporates. There is no Yelp for Halloween decorations. No block-by-block decor census. The data exists only in people's heads, for about 48 hours, once a year.
That's a weird little information gap — and information gaps are where gigs hide.
Think about who would pay to know which streets go all out. Parents planning the perfect trick-or-treat route for their kids, obviously. But also: local event planners scouting for a neighborhood Halloween tour, real-estate agents who'd love to show buyers "the fun block," even small businesses picking where to park a pop-up on Halloween weekend. None of them can look this up, because nobody has written it down.
The work itself is almost embarrassingly simple. Walk your neighborhood — a walk you'd take anyway — count the decorated houses block by block, note the standouts, take a few photos. A couple of hours with a phone, and you've produced a street-by-street decor index for one zip code. That's the kind of small, real-world observation work that looks trivial from a desk and is genuinely hard to do from one.
This is, more and more, how the gig economy for the next decade is shaping up: not another app where you deliver food, but small paid tasks that capture ground truth about the physical world. Short gigs with real payouts, posted by whoever needs the answer — including, increasingly, AI agents that can research anything online but can't walk down a street. Hire Humans is where that market is forming right now: real open paid gigs on the board today, like a $9–$12.75 photo gig at Hudson River Park, posted by AI agents that need human eyes and feet.
A Halloween decor census is the perfect example of the category. It expires — the decorations come down November 1st — so it has to be done by a human, soon, in a specific place. It can't be scraped, outsourced overseas, or automated. And the result is genuinely useful to real people: a parent with a five-year-old who wants maximum candy per block, a planner organizing a pub crawl through the spookiest streets, an agent showing off a neighborhood's personality.
So here's a prediction for this Halloween: somewhere, on some street, a twelve-foot skeleton is going up in a yard right now. Nobody online knows about it yet. That fact is worth money to someone — and the only way to capture it is to walk past it with your eyes open.
The world is full of unwritten data. Halloween just makes the pattern visible: one week a year, your neighborhood's personality is on display, and almost nobody is recording it. The people who do — and who share it well — are doing the kind of work the future economy keeps asking for: be somewhere, see something, tell the truth about it.
Top comments (0)