News & Views

Failure Is Fine As Long As It's Fast / AT&T Foundry

by Contagious Team


AT&T's VP of Ecosystem & Innovation talks to Contagious about how the company breeds successful innovation in its Foundry

Innovation labs are certainly de rigueur, with companies large and small opening in-house R&D branches to get ahead. At AT&T, the so-called ROI of R&D isn't exactly news. The US-based telco's in-house R&D dates back to the creation of Bell Labs in1925 to 'overcome the technological obstacles that were certain to arise in this new industry,' according to a history on the company's website. The ahead-of-their-time labs more than made good on that promise: Researchers from Bell Labs and AT&T Labs (Bell's successor after a 1996 spinoff) have won or shared seven Nobel Prizes in Physics

'AT&T built a solution that was very much purpose-built,' says Abhi Ingle, Vice President of Ecosystem & Innovation at AT&T. According to Ingle, that same solution is no longer sufficient for a large company that wants to compete. He oversees the company's newest solution and AT&T Labs' upstart cousin, AT&T Foundry. Launched in 2011 in Palo Alto, Texas, and California, Foundry's three innovations centers represent $80m of investment on the part of AT&T. The goal of the centers? To deliver customer-facing applications and services at a rapid pace, developed by internal and external collaborators using AT&T APIs specifically designed for developer use.

Ingle's enthusiasm is infectious as he explains the need for AT&T to supplement its traditional research with collaborative innovation. 'If you look at the world and what has happened with technology, there's a massive unleashing of mobile computing technologies happening right now. Innovation can come from any of these areas. We've got to figure out some way to bring some of that innovation in.'





Find out more about how companies are responding to change during Now / Next / Why 2013's Adaptive Innovation segment.

At the heart of Foundry is TIP, short for The Innovation Pipeline, an idea stock market of sorts. Market participants critique and vote for the best concepts submitted to TIP, giving visibility and kudos  to the people with the best ideas. AT&T hoped to get a thousand of its 250,000 employees onto the TIP platform; it got 140,000. 'You can have a frontline employee who is really engaged, with great ideas, who has ten times the voting power of a VP,' says Ingle. 'That is the way it should be. Because it's an idea marketplace, not a hierarchy marketplace.'

That spirit extends to the physical Foundry centers themselves, where internal and external collaborators work on making TIP-selected proposals into reality. 'The culture is very much "Leave your badge at the door,"' says Ingle, pointing outside the building. 'When you're in here you're working on a Foundry project. I don't care what company you're from. I don't care what department you're from. I don't care what your rank is. We are here to do a project. And you're here to innovate.'

Since its inception two years ago, Foundry has produced projects like AT&T mHealthAT&T Messages, and beta video project SimpleView. But Ingle stresses that innovation can't thrive in an environment that only focuses on homerun successes.

'We've had some huge hits out here, and we've had a lot of failures, too. I'm proud of that frankly. If we aren't failing, we aren't trying,' he says, speaking to Contagious inNew York City at a showcase of current Foundry projects under development. Staged inside AT&T's mammoth Long Lines Building in downtown Manhattan, the showcase resembled a telco science fair - complete with poster boards and participants eager to explain their experiments - with smartphones in place of vinegar volcanoes.

Experiments on display at the event included a real-time speech-to-speech translator, a video-based ambient communication system, and an intuitive two-step verification platform. Also present was the Good Times app, which won AT&T's CES Mobile App Hackathon. Good Times uses a brain-monitoring headset and AT&T's call management and machine-to-machine APIs to automatically re-route phone calls based on the mental status of the wearer.

Not all of the projects that come out of Foundry are worth advancing beyond the prototype stage. But rather than judging Foundry's success on a project-by-project basis, AT&T views the centers like a venture capital portfolio, where one money-making project can support many more 'failed' experiments. That position creates an environment that encourages exploration.

Ingle stresses that for innovation to succeed, some structure is needed. At AT&T, that structure comes via a constrained timeline, with Foundry requiring that projects move rapidly. The centers limit experiments - or discrete chunks of a larger project - to a twelve-to-sixteen week deadline. That sort of accelerated approach is what allows the company to embrace failure as a necessary part of innovation.

'A lot of companies (us included) have these very long, large-scale projects where you go into this cave and you isolate yourself from what's happening around you. It's too late to turn around because too much money is already spent. We're trying to avoid that,' says Ingle. 'Two or three huge honking projects that fail quickly instead of failing painfully slowly, catastrophically over time - that by itself pays off the money.'

Sped-up timelines don't only beget failures. One Foundry project, an integrated system called Remote Patient Monitoring built on top of mHealth, succeeded because it iterated quickly on the cheap. Rather than spending months or years analyzing the system's business case, AT&T gave RPM the Foundry go-ahead and the project was up and running within weeks.

AT&T Foundry's approach is reminiscent of Ray Kinsella's vision in Field of Dreams: 'If you build it, he will come.' By creating innovation centers that invite both internal and external collaborators to innovate using AT&T's developer APIs, the company has sparked a new wave of innovation in the company, a light and quick layer of development that complements the ongoing deep research at AT&T Labs.

'It's avoided cost. It's serendipitous discoveries. It's about employee engagement because you release people's creativity. It's about reaching new market segments that you never thought about. It's about monetizing things that are not big enough for you to go into but a developer who chooses to monetize your APIs will go into,' says Ingle excitedly. 'When you put all of these things together, all of a sudden you see this explosion of innovation.'