Joseph Marver sold women’s clothes in Castro Valley, California. October was always dead. Across the street, a costume shop was packed. In 1983 the costume shop moved away, so Marver put his regular stock in storage and filled his store with Halloween. He went to a trade show in Chicago and bought everything spooky he could find. “His sales went through the roof,” says Lisa Morton, who wrote a history of Halloween. He’d found his season.
The next year he tried a bolder version. He took a short lease on a space in a local mall, eight to 10 weeks, and sold costumes, decorations and face paint until the season ended. That was an odd ask at the time. Landlords signed ten-year leases as a rule, says Barbara Kahn of Wharton, and a newcomer had trouble getting any lease at all. Some strip centers were holding out for big-box tenants. When those didn’t come, the malls said yes to a pop-up. Sixteen years on, he had 60 of these seasonal shops under a new name, Spirit Halloween. The off-season was the real work. The team sourced costumes and scouted the next year’s real estate.
Spencer Gifts bought the company in 1999. Marver consulted for a while, then went into the hotel business. The chain he started now opens about 1,500 seasonal stores each fall in the U.S. and Canada. “I didn’t invent temporary sales, but I feel like I invented temporary Halloween,” he told the Seattle Times in 2022. He died in 2025.
Chicken feed and $4 monsters
Candy corn is older and stranger. Wunderle Candy of Philadelphia made it in the 1880s, when nature-themed sweets were the fad. A worker named George Renninger shaped sugar like corn kernels and poured the colors in layers by hand. It sold as a joke version of chicken feed, a penny candy for any time of year. One early Wunderle employee was Milton Hershey, years before the chocolate. Another firm, Goelitz, scaled the candy up and tied it to Halloween. Goelitz became Jelly Belly in 2001 and sold to Ferrara in 2023. Ben Cooper did the same trick for costumes. In 1937 he bought a company that held the Disney costume license. At its height his firm controlled 80% of licensed characters, says a filmmaker who made a documentary about him. The costumes were a plastic mask and a tarp for a few dollars. The company went bankrupt in 1992. And Walter Knott, of Knott’s Berry Farm, built the first big commercial haunted house in 1973 to fill a slow season. Admission was $4. The ads promised live monsters. It’s open 26 nights this year.
Our read
This is a Career story, and it’s about the cheapest way to test a business. Marver didn’t quit, borrow big or sign a 10-year lease. He put his regular stock in storage and tried Halloween for one season. Then he rented a mall space for eight to 10 weeks. Each step was small enough to walk back. Knott’s did the same with a slow season and a $4 ticket. The pattern repeats in every one of these stories. Find the month that’s busy across the street. Run a short, cheap test. Keep the day job until the test pays. No ticker here. None of these companies sits in our books.
The concrete thing in your plan is the side-income line, the one that says what a second business must earn before it changes anything else. If you’ve got an idea, write down what an eight-to-10-week test would cost and what it would have to sell. This week, price that test. Keep it small enough that a flop is a story and not a setback. We’d rather see the test priced before any lease gets signed. That’s what a review is for.