Actors Mark Consuelos and Yara Martinez film a scene from the Amazon series 'Alpha House.' Philip Montgomery for The Wall Street Journal
In May, a dozen Amazon.com Inc. executives, including Chief Executive Jeff Bezos, gathered in a Seattle conference room to select the first original TV shows the company would produce for its streaming video service.
A group of 14 "pilot" episodes had been posted on the company's website a month earlier, where they were viewed by more than one million people. After monitoring viewing patterns and comments on the site, Amazon produced about 20 pages of data detailing, among other things, how much a pilot was viewed, how many users gave it a 5-star rating and how many shared it with friends.
Those findings helped the executives pick the first five pilots—winnowed down from an original pool of thousands of show ideas—that would be turned into series. The first will debut this month: "Alpha House," a political comedy about four politicians who live together, written by Doonesbury comic strip creator Garry Trudeau.
The Internet retail giant, known for efficiency and top-notch logistics, is venturing into the unpredictable game of searching for TV hits. The company is betting it can improve on the traditional TV development process by collecting viewer feedback in unprecedented ways and using it to make less risky bets on which shows to produce. The current system, Amazon says, relies too heavily on studio tastemakers to decide what shows get made.
"We've always operated in a way where we let the data drive what to put in front of customers," says Bill Carr, Amazon's vice president of digital video and music. "We don't have tastemakers deciding what our customers should read, listen to and watch."
By developing original TV programming, Amazon is taking a page from Netflix Inc., which has already had success with shows like Emmy-winning Washington drama "House of Cards." Both companies are leaving their stamp on the industry in different ways. Netflix, for example, doesn't even require pilots.
Amazon's approach could hold lessons for established media companies, who do research on shows before they air but not on anywhere near the scale as Amazon. If Amazon turns projects other networks have passed on into successful shows or discovers untapped talent, Hollywood will take notice. Its efforts are already provoking debate among TV executives about how much of a role big data should play in an industry where creative instincts have always been paramount.
Amazon offers a streaming video service that comes with a subscription to its "Prime" two-day free shipping program, which costs $79 a year. Like Netflix and Hulu, the company has built an extensive library of TV and movies. But ultimately it determined that original programming was critical to attract and retain subscribers.
"We thought logically what we need to have is our own TV shows that give customers a reason why they want to tune in—why they should subscribe," says Mr. Carr.
Cantor Fitzgerald analyst Youssef Squali estimates that Amazon will invest about $1 billion in content this year, with original shows making up less than 10% of the total. Netflix will spend about $2.5 billion on content, Mr. Squali estimates. Since Amazon ramped up spending on streaming video, along with other initiatives like grocery delivery and mobile devices, its profit margins have suffered. The company has reported three quarterly losses in the past year.
"Amazon has always been running on extremely thin margins," Mr. Squali says. The video service, he adds, could become a profit-driver with more scale and if the company charges separately for it in a few years.
The company launched Amazon Studios in 2010. To run it, Amazon tapped Roy Price, an executive with long family ties to Hollywood. The 46-year-old's grandfather was a television producer at Warner Bros. in the 1950s. His father, Frank Price, was a top studio executive at Universal and then at Columbia Pictures, where he oversaw movie releases including "Tootsie" and "Gandhi."
Roy Price is director of Amazon Studios. Philip Montgomery for The Wall Street Journal
Growing up in Beverly Hills, Calif., Mr. Price says he spent time on the Universal lot, chatting with the gate guard, Scotty, and getting swimming lessons from Lee Majors, star of the TV series "The Six Million Dollar Man." After college he spent several years at Walt Disney Co. developing animated series. But he gravitated to the business side of media too: he was a consultant at McKinsey & Co. before Amazon recruited him in 2004 to be employee No. 1 in their nascent digital video business.
Amazon Studios ramped up in earnest this year when the pilots were posted online in April. Users were able to rate pilots with up to five stars and offer comments, just as they would for a book or a George Foreman grill on Amazon's site.
The pilot data was sliced in various ways—the percentage of five-star ratings, for example. Users also could fill out written surveys. Executives in May reviewed data for each pilot as well as recommendations from Amazon's programming team.
Traditional networks test shows, too, but on a much smaller scale—typically focus groups of about 50 people. Amazon is testing on a vastly bigger audience and is collecting a range of metrics unique to its service, such as whether members of its Prime service liked particular shows.
Data wasn't the only consideration. Mr. Price says Amazon executives look at many of the same factors as traditional executives, like the sustainability of the story premise and the ability of the writing team to execute.
"It is not the case that you walk in on Tuesday morning and the computer tells you, 'Doot doot doot, you should order these two shows,'" he says.
Some TV writers are wary of letting data and consumer feedback play too big a role in the creative process. Rhett Reese, one of the writers of "Zombieland," a show Amazon decided not to turn into a series, says most of the customer feedback displayed publicly on Amazon's site was positive but "there was a vocal minority of haters" who had an impact.
"The Internet can be a poisonous place," he says. "Art shouldn't be crowd- sourced."
Even so, he says Amazon told him they passed on his show—a TV remake of a successful movie—partly because the young audience it appealed to was unlikely to pay for the Amazon Prime offering. Despite the outcome, he says he would happily work with Amazon again.
Mr. Carr of Amazon says such demographics weren't a factor in decisions over any show, and that Amazon offers an array of content which appeals to young audiences.
"Onion News Empire," a behind-the-scenes look at a fake newsroom based on the satire of The Onion, was one of the shows on the bubble. Amazon didn't pick up the pilot for series production but is working with the writers and has ordered a few more scripts. One of the writers, Dan Mirk, says Amazon users provided valuable feedback; some said that the show was stuffed with too many jokes for something with the feel of a drama.
"When you go through the normal TV system, you're at the whim of a handful of network execs," he said. "To be able to put our show out there and have everyone see it and respond was great."
Netflix also has challenged Hollywood's traditional approach, and uses its trove of viewing data to help decide which new shows to bid on.
“Writers and producers say Amazon and Netflix offer far more creative freedom than traditional networks.”
TV writers and producers say both Amazon and Netflix offer far more creative freedom than traditional network executives, who are known for requesting changes on everything from a joke's punch line to a scene that's too gory.
"So many notes at networks are fear based," says Jenji Kohan, a veteran TV writer who created Netflix's prison comedy "Orange is the New Black." "They're worried—how are people going to react to this?"
Netflix does weigh in sometimes. On "Orange is the New Black," the company wanted extensive composed music, so that viewers bingeing on multiple episodes wouldn't hear the same riffs over and over. "You have the opportunity to think of it as a 13-hour movie and how that would be scored," says Cindy Holland, Netflix's vice president of original content.
A big plus for creators is that Netflix doesn't require pilots. Ms. Kohan of "Orange is the New Black" says that was hugely appealing since pilots, designed to show potential buyers of programs how characters and story arcs might develop, can be wasteful and creatively stunting. "I don't want to make a sales tool, I want to make a show," she says.
Seasoned TV executives say networks have valuable expertise that helps show creators develop their ideas and build audiences. Kevin Reilly, entertainment head for the Fox broadcast network, says he's open to experimentation outside the traditional pilot process—Fox recently ordered ancient Egypt drama "Hieroglyph" straight into series production without a pilot requirement. But there need to be some "checkpoints" for creators, he says. "Even the most talented people sometimes get off on the wrong track—and it's hard to change the wheels at 60 miles an hour."
To attract ideas from writers without deep Hollywood connections, Amazon set up an online portal to collect scripts. Among the more than 5,000 submissions, which vary greatly in quality and sophistication, are a parody cartoon about the U.S. Founding Fathers, a comedy about men who start a baby-sitting franchise and a show about a man whose grandfather rises from the dead to give him advice.
"Think about all the ideas that don't become pilot scripts," says Mr. Price, director of Amazon Studios. "One wonders how many other great shows are out there."
For the most part, Amazon's show ideas are coming in the old-fashioned way: through in-person connections with Hollywood agents and established creators. So far, two of the online submissions have made it through to the pilot stage.
One that made the cut is "Gortimer Gibbon's Life on Normal Street," a live-action show about the adventures of quirky 11-year-olds. The writer, David Anaxagoras, is a full-time preschool teacher who struggled for years to sell movie screenplays in Hollywood.
"Amazon has lowered the barriers of getting a script in the right hands," he says. "I don't think anyone else would have bought this." He feels that portrayals of kids on TV can be too simplistic—they're often very happy—and he wanted to explore a more complex set of emotions.
Amazon pays $55,000 for scripts submitted online. If the pilot is successful and the series goes into production, creators get up to 5% of merchandising receipts and a per-episode fee of $4,000 for a one-hour show and $2,500 for a half-hour show. Deals set up offline with more established creators vary depending on the writer's reputation. One uncertainty for Amazon creators is how much revenue potential there is from "back end" proceeds such as syndication and DVD sales.
On a recent afternoon, Amazon's Mr. Price visited Kaufman Astoria Studios in Queens for an "Alpha House" shoot. The crew was filming a scene in which actor Mark Consuelos, who plays a freshman Republican senator with presidential aspirations, was practicing his rebuttal to a speech by President Barack Obama. "The president wants to let millions of illegals step out of the shadows," his character said. "I say they should step halfway out of the shadows—just enough so the light hits their faces and we can identify them."
"Alpha House" is based on the real-life story of four U.S. members of Congress who bunked together in a Washington, D.C., townhouse.
Though Amazon's selection process was unique, in many ways "Alpha House" is like any other slick TV production. Big-name actors are involved, like John Goodman, who plays one of the four politicians.
The healthy budget, on par with cable networks, is evident from the set, which has a mock-up of a U.S. Senate building so detailed that it bears the correct door textures and room placards. The office of Mr. Goodman's character has a framed letter from President Ronald Reagan: "Thank you for your strong support in passing the Nuclear Waste Policy Act."
"Alpha House" scored well in the viewing data and got a strong average rating of 4.1. The company also took into account that the work of Mr. Goodman was popular on its service.
In coming weeks, Amazon will find out if it can build a sustained audience for the show. Actors involved say that despite the company's unusual development process, they believe the material being produced rivals traditional networks.
Mr. Consuelos says he believes the quality of his Amazon show is on par with other projects he has worked on for cable networks, despite its unconventional development. "It's a whole new frontier," he says.
Write to Amol Sharma at amol.sharma@wsj.com