Inside Chromecast

The untold story of Google’s revolutionary dongle — and how it changed innovation at one of the world’s largest manufacturers.

Two years ago Google tried something unusual. Rather than handing its manufacturer the mostly baked designs for a new device, it shared the earliest concepts of the product and asked for help designing, engineering, and manufacturing the device at scale. The then-secret gadget would allow users to instantly “cast” content from mobile devices or laptops to home televisions, the way similar products did at the time.

But instead of using clunky boxes like the others in the streaming video market, the Chromecast device would be sleek, curved, and about the size of a thumb drive (2.83 inches in the end). Its wireless technology would have to be crammed into that tiny, unique form. While competitors offered their boxes for upwards of $100, Google wanted Chromecast to go for about $35. Oh, and to get a jump on the market, Google wanted a prototype made in four weeks.

“We didn’t even know how to get to the price point when we agreed to do it for them — we actually just took a leap of faith and said ‘We’ll go figure it out,’” says Mike Dennison, president of the Consumer Technologies Group at Flex, who works out of the company’s California innovation lab. “We were at twice the price when we first started designing this thing.”

Until this point, Flex, a 30-year-old Fortune Global 500 contract manufacturer with a workforce of 200,000 employees operating in more than 30 countries, typically worked off of detailed design plans provided by its customers. Mostly its mission had involved making goods at scale, as efficiently as possible. But Google’s streaming-stick ambition and vision pushed both companies into new territory. The Chromecast symbolized a new way of working that would permeate Flex’s entire operation and blaze a trail for others in the booming business of smart, connected things. For both companies, the project was an experiment inco-innovation, a marriage of Google’s disruptive ideas with Flex’s global design, engineering, manufacturing, and marketing know-how.

There were a lot of intense moments. There were multiple prototypes that we built. Early on, our first pass didn’t actually work, Kapur says.

“It was an opportunity for us to really pioneer a new space, where we were taking something from sketch to concept to consumption,” Dennison says.

First, however, Flex would have to leap a series of technological hurdles and meet near-impossible deadlines.

Few engineering or design challenges really concern Dr. Joan Vrtis, chief technology officer at printed circuit supplier Multek, a wholly owned subsidiary of Flex. Vrtis and her team work across aerospace, medical, automotive, consumer, and other industries. They make the paper-thin circuit boards found in the sleekest electronics in wearables, homes, and more. But when it came to Chromecast, named after Google’s leading web browser, Chrome, material utilization and efficiency were her concerns. Printed-circuit engineers lay out circuitry to fit in square or rectangular shapes. It’s how they get the most out of printed circuit material and control costs. “Every process in our factory uses the same panel size to ensure efficiency,” Vrtis says. “We use lasers and routers to make straight cuts that separate printed circuit boards from these panels.” However, Google’s Chromecast required a circular printed circuit board, which cost 20–25 cents and dropped the panel layout efficiency as much as 24%.

First, Vrtis sought to streamline the process while satisfying Google’s need for a unique form factor. “Then I’m thinking ‘How do I discourage them?’” she says. “Even the guy at Google who knew printed circuit boards looked at me knowingly and said, ‘This is going to be a challenge.’”

Besides price, Google had to get the Chromecast to market with unprecedented speed — there was a significant opportunity to seize market share in the fast-growing sector of home electronics — called over-the-top (OTT) video streaming services — with a small and inexpensive device that was easy for even tech-averse consumers to use.

It was an opportunity for us to really pioneer a new space, where we were taking something from sketch to concept to consumption, says Dennison.

“Timeline was a significant challenge,” says Sumir Kapur, Flex senior vice president of global accounts who manages the Google partnership. And in the four weeks the company had to develop the prototype, “There were a lot of intense moments. There were multiple prototypes that we built. There were four to five iterations. Early on, our first pass didn’t actually work. We had to go back and essentially start from scratch,” he says.

The wireless aspect turned out to be the greatest challenge. On top of getting a signal to a device nestled in between the back of a TV and a wall, “How do you lay an antenna in there in that design?” Dennison recalls wondering. “When you’re in this mode, you’re literally trying everything under the sun, from an antenna that sticks out to an antenna that’s on the board to an antenna that’s wrapped around it.”

Then heat became an issue: How would they keep all of the circuitry crammed into the tiny device from melting its circuit board or shell? “If you use heat shields on top of an antenna in that space, you’re in trouble,” Dennison says. “We tried different things, from venting it differently to mechanical attributes to get more air into it, but space was always an issue. And height. We didn’t want it to be any thicker than a USB. But you could have the coolest iconic device in the world; if it doesn’t work, it’s a paperweight.”

The solution was no singular breakthrough. It was a jigsaw puzzle: an octagonally shaped printed circuit board that took advantage of the gaps between circuits to disperse heat so it wouldn’t toast the circuitry; and an antenna developed with Multek that was a ribbon nestled between the layers of the circuit board that could sniff out a signal from behind a flat-screen.

Chromecast was one of the first times we didn’t have the clear recipe in front of us. We had to iterate toward what Google ultimately wanted in its initial concept. It’s what we call innovating on the factory floor, Dennison says.

To be clear, the Chromecast isn’t quite one of the intelligent things Flex is known for designing and manufacturing today — from ski boots that help skiers correct their form by tracking the movement of their feet, to coffee cups that warm themselves back up after the brew gets cold. These are the devices that, in the near future, will know when you enter the living room, as do the smart light-bulb sockets that Flex is gearing up to produce with one of its startup partners. In the coming age of intelligence, screens might anticipate when viewers want to change the channel. Devices might connect with smart clothing that tracks responses to video programs by measuring biorhythms through embedded sensors. Chromecast wouldn’t be that intelligent in its initial version. But it was a design, engineering, and strategic triumph. It would bring online media to an unprecedented mass market — serving as a sort of Trojan horse for Google and its YouTube video franchise. As Wired’s Mat Honan asked and answered in an early review, “Did Google just win the living room? No. But it definitely just cozied up on a lot of couches and got comfortable.”

The stick’s signature size and curves, meanwhile, appear to have been worth the extra effort, visually distinguishing the Chromecast from both its competitors as well as from the growing sea of rectangular USB-port plug-ins that many consumers amass in their living rooms.

Google began selling its Chromecast streaming sticks for $35 a pop in July 2013, and Flex says it cranked out 7 million of the devices during the first four months of production, handily beating Google’s competitors to the streaming-stick market while offering a steeply discounted alternative to Apple’s $99 Apple TV set-top box. The Chromecast has overtaken Apple TV, in fact, as one of the most popular Internet video streaming devices in the U.S., according to a survey research firm Parks Associates released in December. It was Amazon’s best-selling electronic device as of this fall, when the e-commerce giant decided to launch its own Fire TV Stick, almost a year and a half after Chromecast went on sale.

“There has been no product that I know of that had such a short development cycle that has taken off in the market quite like this one,” says Kapur.

To be sure, Amazon’s $39 Fire TV Stick boasts more storage and memory along with a different user interface, while the $50 Roku Streaming Stick, the latest version of which was introduced in early 2014, supports more apps than the Chromecast and, by some accounts, makes searching for content easier. But no matter how reviewers compare Chromecast to newer streaming gadgets, they rarely fail to mention the device as the standard-bearer. It’s a testament to the importance of speed to market, cost, and a determined vision — made real at scale by Flex’s innovation team. “I think it was a collaborative approach that yielded a product that performed extremely well at a really low price point,” Kapur says. “We and Google together established the market for this product.”

Engineers such as Sinat Le (pictured) at work in one of Flex’s labs of the future, which incorporate groundbreaking manufacturing technology. Photos: Patrick Molnar

In the same way Flex called upon its experience with other technologies to crack the circular Chromecast conundrum, it’s applied the lessons of that device to its core business strategy. Today, it’s grown from a traditional manufacturing powerhouse to one of the biggest and most popular co-innovators in Silicon Valley.

“That shift has been huge for us. Chromecast was literally the pioneer of that way of thinking,” Dennison says. Part of the reason Dennison thinks that companies such as Google are now looking to Flex earlier than ever in the product development process is to “solve the problems of innovation.” In an age of disruption, where speed to market makes the difference between pioneers and also-rans, innovation has to happen on parallel paths, and few innovators have the ability to go it alone.

“We’re seeing invention, or innovation, happening at the same rate as productization,” Dennison says. “You’re literally inventing components and concepts and sensor technologies at the same rate that you’re developing an end-state product. Chromecast was one of the first times we didn’t have the clear recipe in front of us. We had to iterate toward what Google ultimately wanted in its initial concept. It’s what we call innovating on the factory floor.”

Today at Flex, “innovation has to be the centerpiece,” says Flex Chief Executive Officer Mike McNamara, who has led the company’s evolution for 20 years. For the first 10 of those, the business was all about “labor arbitrage,” he recalls — helping customers manufacture various pieces of their products in markets where labor was cheapest, often in China or Mexico. But as labor costs overseas have crept up, reducing such arbitrage opportunities, Flex is increasingly luring customers instead with a growing arsenal of technologies that it has developed to accommodate the biggest companies in nearly every industry — proven building blocks that Flex can use to speed ideas to market. And thanks to its uniquely omniscient position in the secretive Silicon Valley, Flex can transplant ideas from one industry to another. Ideas from completely different (and always non-competing companies) are part of an institutional cross-pollination that allows Flex’s customers to get to market quicker and compete, McNamara says. “Cross-pollination is where Flex lives.”

Recently, for example, Flex borrowed a server-cooling technology from its computer division and used it as a low-cost way to cool down powerful warehouse light bulbs, “technology that nobody who is only in the lighting business would think to use,” McNamara says.

The company is repurposing an electro-adhesion technology, which is being developed by a startup partner for factory assembly lines, to help vacuum maker Dyson create its next generation of dust-busters.

With Google, the takeaway wasn’t as much about sharing a specific technology breakthrough as it was about proving the concept of co-innovation, beginning at a product’s earliest stages. Designing, engineering, and manufacturing hardware at the same rate that companies such as Dyson or Google are developing or integrating software technology is incredibly hard, Dennison says, but in the end,

“What we’ve created was better than what we all intended.”


If you enjoyed this article and would like to read more, please check out

INTELLIGENCE explores the concept of co-innovation and the “Intelligence of Things,” that Flex sees as the building blocks of the post-Information Age era. More at

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.