Ian Henderson, 26, of Germantown, Md., right, chats with a customer after returning the man's Kia sedan to him last year in Washington. (Faiz Siddiqui/The Washington Post)

It seemed almost too good to be true: $200 a month for unlimited parking in Penn Quarter. No need to navigate the constellation of signboards outside parking garages; prolonged morning searches for a spot would become a thing of the past.

Kate Genser, 35, of Petworth said Zirx on-demand valet was ideal for parking at her communications job near CityCenterDC. With a tap on a smartphone icon, she would summon an attendant who would arrive via scooter. She would hand over the keys to her Nissan Versa and head into work.

“I used it as my commuter pass,” she said. “My logic was if my garage is $13 a day, that would be $260 a month. Sixty bucks of savings — why not try it?”

But a year after it launched in the District, on-demand valet parking has proven unprofitable and difficult to sustain. A month after her January monthly pass went into effect, Genser received an email from Zirx that said, “Unfortunately, we are temporarily pausing our service in D.C.”

That left the nation’s capital devoid of an industry that investors had hailed as the future of parking — the “Uber of parking,” if you will — and a lifesaver for motorists in cities such as D.C. where spaces are scarce and expensive.

Ian Henderson, 26, of Germantown, Md., races between downtown parking garages last year as a valet for Zirx in Washington. (Faiz Siddiqui/The Washington Post)

In San Francisco, for example, attendants for Luxe, wearing blue jackets, competed against yellow-clad Zirx valets. For a brief period, pink-clad attendants parked for Carbon, another on-demand valet service that has since folded.

But the idea never really caught on in the District or some of the companies’ other markets.

“I mean, we were wrong,” said investor Ethan Kurzweil, who was involved in a multimillion-dollar investment in Zirx from Bessemer Venture Partners. “Economically, on-demand parking didn’t work.”

Zirx, which was the first such service to launch in D.C., halted service earlier this year in all six cities in which it operated: San Francisco, Seattle, Los Angeles, San Diego, New York and the District.

That leaves Luxe as the major player in on-demand valet. But Luxe has no plans for expansion at the moment, a company spokesman said, and intends to focus on its six existing markets: San Francisco, Los Angeles, Chicago, Seattle, Austin and New York. Late last year, it pulled out of Boston and Philadelphia. The company said it wants to aggressively home in on the user experience by narrowing its coverage area.

Two of its six markets are profitable today, including New York, the company said.

“D.C. is definitely on the next set of cities that we’re evaluating. We were thinking about launching there last year, and we decided not to,” said Luxe chief executive Curtis Lee.

“We love D.C., and D.C.’s a city that’s perfect for this kind of service,” Lee said.

For Zirx, the on-demand model simply wasn’t turning a profit, said CEO Sean Behr. It was too hard to flood the streets with enough valets to keep wait times down while also making money.

“The number one reason why I believe that on-demand valet parking doesn’t work as well as I’d like it to work is the nature of the supply,” Behr said. “I would argue that the supply of parking spaces is fixed and inelastic. And in on-demand companies, what you really want is variable and unlimited.”

This month, the company announced it had shifted resources to serve as a software platform for businesses — a more fruitful model, Behr says.

Its new customers include Openbay, a mobile platform for arranging auto repairs, and Bama Commercial Leasing, which leases cars to some Uber drivers. The existing network of valets can transport cars and provide concierge services — oil changes and gas refills — for companies, who pre-arrange the jobs, Behr says.

The refocused company expects to break even in all eight cities where it operates by next month. That includes the six where it previously offered on-demand valet, plus Boston and Chicago. That’s in contrast to January, when, Behr says, it was losing money in nearly all of them.

Because Zirx had to prepay for its parking, unfilled spots meant lost money. Turnover was another issue: Ideally, the same spot would be used at different times by three or four cars per day, Behr said, but that rarely proved to be true.

And the parking spots, leased at peak hours in places such as Dupont Circle and Farragut Square, didn’t come cheap.

“If you look at most parking lots, they’re generally full during the day Monday through Friday and pretty empty on the weekends and at night,” Behr said. “That’s the problem. When you really need the parking inventory is when the parking lots are mostly full.”

Companies had poured more than $36 million into Zirx, which secured funding from Bessemer Venture Partners, BMW iVentures and others. About 400 drivers nationwide worked for the company as paid contractors making an hourly wage. Drivers earned bonuses depending on customer volume, company officials said.

Toward the end of January, as Zirx moved toward halting service, Genser said she noticed that wait times skyrocketed.

One night, while dining with friends at a restaurant on 14th Street, she summoned a valet to return her car from Penn Quarter at about 8:30 p.m. She was shocked to see a wait time of 45 minutes pop up on her phone. Normally, it’d be five or 10 minutes.

When her friends left the restaurant, she browsed the nearby Trader Joe’s to pass the time.

Eventually, she said, “I called them, and I was like, ‘This can’t happen.’ The guy hustled over to my car — it was very sweet.”

She said she often felt bad for the attendants, who worked in rain, sleet and snow, and for the company itself, which she thought couldn’t possibly be turning a profit.

“In my mind, I couldn’t make the math work, and I felt bad because I knew they could not possibly be making money on my $200 a month,” Genser said.

Kurzweil said it became evident that Zirx’s business model was flawed.

“Uber’s a better business than this turned out to be,” Kurzweil said. “Uber only has drivers. Uber is such that you can charge enough to pay people something they will provide the service for.”

Behr, who described on-demand valet as like “magic” in a previous interview last year, said he believes in the product but that someone else will have to make it work.

The most likely candidate appears to be Luxe, which announced this month that it had secured $50 million in funding led by Hertz Global Holdings, positioning it as the industry leader and more than doubling Zirx’s investments. Lee declined to go into specifics about the company’s profits.

Behr, who says on-demand valet is “near and dear” to his heart, wished his onetime competitor luck.

“I think figuring out how to get that model to work will take time and will take money,” he said. “I hope that they’ll be able to get it, figure it out and it will work.”