PALO ALTO, Calif. -- Just a short bicycle ride from Stanford University, there's a work space befitting any hip startup.
Squat, teardrop-shaped chairs and short, lime-green swivel chairs sit atop a colorful checkered carpet, across which a small dog casually pitter-patters. Lining the sparse ceiling are rows of florescent lights, wooden beams, and large ventilation tubes, adding an industrial-chic ambiance. On one side lies a sliding divider made up of strands of cable lines strung closely next to each other, while the other side features a movable row of hanging chains, creating a flexible space that can be manipulated as needed.
Aside from a corner filled with a few developers quietly typing away at their computers, it's fairly empty in the middle of the afternoon. For most folks who work here, the afternoon is still early; people tend to flock in at unusual times, and all-nighters are a common routine. A week later, the place will be crammed with nearly 200 developers participating in a Facebook hackathon.
The office, a stone's throw from the humble beginnings of startups such as Flipboard, is a magnet for venture capitalists and developers -- Marc Andreessen is a familiar face. But it's no pre-IPO startup. It's AT&T's application foundry, where the company has created a place and provided the resources for local developers to work on small projects with the potential to change the multibillion-dollar telecommunications giant.
What AT&T has been doing over the past few years is jump-starting a culture of innovation within the company. AT&T, far removed from its roots as a pioneer in communications technology, has spent the last few years bringing back some of that fearlessness in experimenting with new projects -- crucial at a time when all the tech heavy-hitters have shown a willingness to try new things.
At the heart of AT&T's innovation drive is its technology council, responsible for the creation of the foundry and several other key projects within the company. It's a handpicked brain trust that's helped the company shed some of the bad habits of a large corporation.
"We take on problems that (AT&T) inherently wasn't going to address well because of our size," John Donovan, senior executive vice president of technology and networks for AT&T, said in an interview with CNET.
The technology council is antithetical to the popular perception that AT&T is a slow and lumbering company unable to get even its cell phone reception issues under control. The group has been able to spark a legion of ideas through a massive crowd-sourcing effort; opened the company up to building more relationships with the startup community; and, yes, even helped to improve the network. Equally important, it's redefined the meaning of speed within the company, and bolstered a willingness to experiment, even if it means failure.
"What it does is provide an organized channel to use the ingenuity of AT&T's employees to bring these ideas quicker to use," said Roger Entner, a consultant with Recon Analytics who follows telecommunications companies.
A change of pace
AT&T's willingness to look outside for help on projects is a change from its original preference to develop new services and products in-house, relying on its once formidable AT&T Labs business. Under Ma Bell, it had a massive research and development arm in what is now the Lucent part of Alcatel-Lucent.
AT&T has poured money into the new initiatives started by Donovan. The company doesn't break out its investment in this area, or in research and development, but it's a drop in the bucket relative to the roughly $20 billion it spends on capital expenditures each year. Still, the foundry has become a vital component of the R&D budget, Donovan said.
AT&T isn't innovating for innovation's sake -- at stake is its identity. If the company doesn't stay in the cutting-edge game, it'll lose its relevancy in a world filled with Apples, Googles, and Facebooks. Worse yet, it risks relegating itself to a dumb-pipe, or a basic connection over which all the lucrative services and applications ride, and which AT&T doesn't get to take part in.
What started as a gradual break from its rigid structure as a traditional phone company has transformed into an intense effort to stay toe-to-toe with the most nimble companies in Silicon Valley.
The push comes as competitors such as Verizon Wireless and Sprint Nextel have similarly abandoned the notion of closing off their businesses, and more closely work with startups and companies they previously would have seen as threats, such as Google and Apple. AT&T's exclusive access to the iPhone, for instance, drove Verizon to embrace Google and heavily push Android smartphones.
AT&T certainly isn't without its share of issues, and its reputation for moving slowly is often justified. The company, for instance, took its time addressing the network problems that cropped up in the early years of the iPhone, wrecking its public image for years. Like the other carriers, it's slow to provide the latest software updates to Android smartphones.
That doesn't discount the early bets AT&T made with initiatives such as its U-Verse Internet-based television service, which suffered through many hiccups in the early years. It was also the first to move into a tiered data pricing structure for wireless plans, which created a lower-cost option for first-time smartphone buyers even as it irked hardcore subscribers who spent a lot of time -- and downloaded a lot of data -- on their mobile devices.
That innovation has fueled its most recent -- and somewhat head-scratching -- move into the home security and automation business.
The smartest minds in one (virtual) room
One of the first things Donovan did when he joined AT&T in 2008 as chief technology officer was to ask the managers in the company to point out the brightest individuals "who weren't afraid to speak the truth." He wanted to bring together a group of people answerable only to him for the purposes of shining a spotlight on the blind spots within the company.
As much as AT&T has loosened up over the years, it's fundamentally still a company in which suits are regularly worn, employees still use courtesy titles, and a strict chain of command is still observed. So it was an understandably nerve-racking experience for the first 16 people who were chosen to meet and serve with the newly hired executive.
Their fears were assuaged the first time they sat down with Donovan. "In the first few meetings, we were sweating bullets until we realized it was John, not Mr. Donovan," said Adam Hersh, who helps commercialize new projects and ventures for AT&T, and was an original member of the tech council.
- AT&T unloads Connecticut wireline biz for $2B
- The prodigal daughter: Can she bring her Verizon LTE smartphone to AT&T?
- iPad Mini availability at Verizon, AT&T improves, supply eases
- Privacy advocates blast AT&T's alleged sale of info to CIA
- New York AG to cellular carriers: Why was antitheft switch killed?
Donovan talked with them about their ideas, suggestions, and complaints. The first session lasted five hours, with the new boss running through every point they had brought up, challenging them and asking several rapid-fire questions. It wasn't the typical company meeting.
The big brains come from all facets of AT&T's technology divisions, including research, labs, engineering, mobility, and intellectual property, and they brought their own unique perspectives. Their responsibilities on the council come on top of their regular duties.
"It's above and beyond your normal job," said Sanjay Macwan, vice president of technology and innovation for AT&T and another original member who now leads the council.
The council members meet at least twice a week. What started as a series of phone conference calls has evolved into regular telepresence calls, and face-to-face meetings every time the members find themselves in the same city. Donovan, who was promoted to his current role in charge of technology and network operations, still meets with the council alongside new Chief Technology Officer Krish Prabhu every six to eight weeks.
The council is seen as a springboard for one's career inside AT&T. Hersh, for instance, worked in network planning and engineering when he was chosen to participate on the council but was promoted to help usher in new projects after his tenure in the group ended. Macwan used to run security and mobility and was promoted to work more actively with the startup community and lead the council.
Council members are rotated in and out to keep the ideas fresh. For every position that's open, there will be 30 to 40 inquiries from interested employees, Macwan said.
"It's a privileged group to be a part of," he said.
The tech council at AT&T isn't entirely unique. Ericsson, for instance, has a similar organization within its company, according to Chief Marketing Officer Arun Bhikshesvaran. But AT&T executives believe that no other company has as dedicated a group devoted solely to coming up with new ideas and projects.
The appeal of serving on the tech council isn't just the implications for one's career; it's also the opportunity to drive some big changes at the company. In its short time, the technology council has been responsible for some major changes and initiatives at AT&T.
In search of TIPs
Donovan liked the idea of drawing on the massive resources within AT&T, so one of the first major projects that emerged out of the technology council was the innovation pipeline, or TIP, a crowd-sourcing machine designed to get more employees involved with change at the company.
What started with a handful of people contributing has grown into a massive internal social network constantly coming up with ideas, providing feedback, and deciding whether those ideas should move on to get funding.
The crowd-source mechanism isn't new. Daimler Chrysler, for instance, has a similar program in place. But there are few as large as what AT&T has managed to put together.
While it started with just the technology group, TIP has been expanded to include every employee of the company. The program has roughly 138,000 AT&T employees involved in some capacity, with 11,000 "founders" who take ownership of an idea, help foster it, make suggestions, or bat down ones that aren't ready for prime time.
There isn't a shortage of suggestions. In fact, 19,000 different ones have been submitted since the program started three years ago, Donovan said. Each quarter, the crowd determines the 10 best suggestions to be presented at a fast-pitch session, where the technology council chooses three for funding from AT&T's internal angel funding mechanism.
Employees who see their projects get implemented are rewarded with trophies, executive recognition, and bonuses.
In one quarter, an intern managed to win the crowd's support, and ultimately get an investment from AT&T, while a senior executive's pitch got batted down by the TIP users.
"We're democratizing the idea," Macwan said. "It stands on merit alone."
Still, there are no penalties for getting a suggestion rejected. Even for the projects that get funded, the company embraces the concept of a fast-fail model; a project has about three months to prove itself before the company shuts it down and moves on. Executives say projects under this program move three times as quickly as the normal AT&T process.
"We're absolutely willing to fail," Macwan said.
The foundry project
The internal fast-pitch model developed through TIP was the basis for a broader set of speed-dating sessions that Donovan wanted to bring to AT&T that would allow it to tap into talent outside of the company.
But rather than just holding a series of meetings, Donovan and the tech council opted for the more ambitious idea of building innovation centers around the world. These weren't extensions of AT&T's Labs unit; they were places where outsiders could come with ideas that could benefit AT&T, and get support, resources, and a base of operation for their work.
AT&T started with its first foundry in Plano, Texas, before moving on to Ra'anana, Israel, and then Palo Alto, Calif. Texas dealt largely with enterprise ideas, Palo Alto with consumer, and Israel with network infrastructure, although ideas and concepts were shared throughout the three facilities.
These offices often lack the polish or scope of the corporate headquarters. The main room of the Palo Alto foundry, for instance, has the decidedly politically incorrect moniker "French Laundry," a 19th century reference to exclusionary white-run laundry businesses and a callback to the space's previous life as a Laundromat. Curiously, it has nothing to do with the exclusive restaurant of the same name in Napa Valley, Calif., which is a favorite haunt of the Silicon Valley elite.
AT&T employees at the foundry embrace the ironic undertone of the name, given the current inclusive nature of the workspace. Networking events and casual social mixers, for example, are regularly held there.
It's a place that brings together people with different technical, business, and creative perspectives without the formal titles and bureaucracy, said Mark Negal, executive director of the foundry program in Palo Alto.
"Our goal is to accelerate work on new services," he said.
Projects at the foundry move three times as fast as initiatives taken on elsewhere at AT&T.
In Israel, Hersh and his colleagues didn't even wait for the facility to be built to do their first fast-pitch session with local startups; he held the initial one on the squash court that eventually became the company's foundry.
When Donovan started at AT&T, he contacted the various major venture capitalists, such as Sequoia Capital, Kleiner Perkins, and Stanford, and met with many startups. With the three foundries in place, he had formal locations to entertain pitches and provide support and resources to the startups that the VCs trotted in.
The company is an active participant in Silicon Valley, said Jim Goetz, a partner at Sequoia. That's a dramatic change from a few years ago.
"Before, you had to fly down to San Antonio or Dallas to visit them, and there were few individuals there who could make decisions without a committee," Goetz said.
Beyond funding new ideas, AT&T has also used the foundry to marry up TIP ideas with third-party developers. An AT&T employee came up with the idea of a smartphone with dual identities -- one for work and another for personal use -- that were kept separate from each other, allowing individuals to make personal calls and e-mail and send text messages to friends and family members without jeopardizing sensitive corporate secrets.
Through a fast-pitch session in Palo Alto last year, AT&T found a company called Enterproid, which developed a similar-sounding app called Divide. The ideas were combined and AT&T unveiled its Toggle service at last year's CTIA Enterprise & Applications conference.
Of course, the ideas don't always pan out. Toggle, for instance, hasn't really caught on with businesses, and critics pointed out some of the limitations and hassles that come from running two separate and distinct profiles on a phone, as well as its initial Android-only access. AT&T yesterday relaunched Toggle for iOS with a new partner, and is working on a version for PCs and Macs.
Still, there are other examples of potential projects, including a connected telemetric device that monitors how fast a car is moving, can deliver "report cards" to parents of young drivers, and is something AT&T is working to bring to market. Others include the release of an application programming interface tool for use with HTML5 Web apps, which was unveiled in January at the company's developer summit in conjunction with the Consumer Electronics Show.
Within AT&T, the foundry concept has become synonymous with speed and how the broader company should be operating.
"'We're going to foundry it,' has become a common expression," Donovan said regarding projects that require tight deadlines.
Earned startup cred
AT&T's foundry initiative in Palo Alto has paid off beyond bringing more innovation in-house. The company's early push to reach out to the startup community has resulted in a lot of credibility and goodwill in the Silicon Valley community.
"If you look at the actual program and level of success, there's no comparison," Goetz said. "The AT&T folks are more visible here in the Valley."
AT&T isn't the only one moving into Silicon Valley, with other carriers such as Verizon, Sprint, T-Mobile USA, as well as international telcos such as Vodafone also planting roots in the region. But AT&T got there first with Donovan laying down the groundwork roughly four years ago.
Generally, it's difficult to get any kind of feedback or follow-up comments from a carrier. But AT&T is easier to work with, responds quickly, and has really improved its outreach program for startups, Goetz said.
"The lines between companies tend to blur as people focus on accomplishing a goal," Nagel said.
AT&T's credibility means it typically gets first dibs on the local talent, particularly as more carriers attempt to emulate its model.
"It's refreshing," he said. "The Valley is often word of mouth. AT&T has real mojo and interest from the best developers."
Jon Auerbach, a general partner at Charles River Ventures, puts AT&T and Verizon on par when it comes to work with startups. He did note that AT&T's work in this area has dramatically improved over the past few years, something he attributes to Donovan's presence.
"John understood the importance of startups," Auerbach said. "Before John came, "AT&T was more bureaucratic."
Prior to Donovan joining AT&T, Auerbach had ranked the company last in terms of innovation and willingness to work with others.
Tapping outside talent
Beyond the foundries, AT&T has been aggressive with its fast-pitch and speed-dating sessions, meeting with companies all around the world. One of the meetings held early last year at its Israel foundry is expected to yield better network performance for AT&T.
When Intucell CEO Rani Wellingstein made his pitch to Donovan directly, he didn't expect AT&T to embrace his company as quickly as it did.
"We were shocked at the speed in which AT&T decided to move and got their act together," Wellingstein told CNET.
Intucell developed a technology that allows mobile operators to fine-tune and adjust their network in real-time based on traffic and potential congestion. Carriers previously would have to look at the data after several weeks and make a decision on whether to invest in more equipment. Intucell's technology meant a more responsive network that delivered a better level of voice and data service.
Wellingstein previously worked at NKO, a pioneer in Internet-protocol phone technology that was acquired by ECI Telecom in 1998. At ECI, he worked with phone product development and dealt with the carriers in North America.
So for a startup like Intucell, he knew the typical route would have been to go after smaller carriers looking to tinker with new technology, before moving on to the big wigs. But after meeting with Donovan in January and getting technically validated in the following few weeks, Intucell had a contract for national coverage signed by November and its initial deployment in December.
"It has to be a record in this industry," Wellingstein said.
AT&T's quick move allowed it to adopt Intucell's technology first in the United States. Wellingstein said he is in discussions with other carriers, including ones in the U.S., and none have moved this quick.
All based on a meeting that lasted less than 10 minutes.
"There is the stereotype that innovation goes to die at big companies," Entner said. "In some cases, that's certainly true. But not in all cases."