It's called the technology petting zoo. Stocked with the latest high-tech gadgets, games, systems and software that could potentially be of business value, it's a place where engineers and other IT users at NASA's Jet Propulsion Laboratory can try out the steady stream of hot, new consumer technologies and imagine the possibilities.
"We can afford to buy at least one of everything that looks like it might have business value. The petting zoo is where they can test it," explains Tom Soderstrom, CTO at the Jet Propulsion Laboratory.
"We also have a social networking site where they rate what they try. It works to create a kind of technology pull, rather than push, and creates a desire among users to work with IT. A big benefit is that users no longer have to be in the shadows and hide what they're trying from the IT security police."
At Pharmaceutical Product Development Inc, CIO Rob Petrie has set up his own version of a tech petting zoo - or skunk works, as such initiatives are often called. The PPD's version is called the innovation group, and it's comprised of a rotating team of IT staffers who come from diverse backgrounds and are taken out of their regular jobs to spend six months experimenting with various technologies. Their goal is to discover ways that new technologies could be applied to deliver business at PPD.
"We always have at least five or six things in motion in the group," says Petrie. "Bringing different people through and then returning them to IT is a way to cross-pollinate ideas. It spreads innovation."
Its name may vary, but there's little doubt about it: the IT skunk works is making a comeback - not so much as a place to allow experimental projects to flourish, away from the stifling corporate bureaucracy, but as an indispensable tool to breed innovation, learn about risk, build employee loyalty, conduct pilot projects and educate technical and nontechnical staffers in a time of rapid-fire technology change.
Some skunk works are formal operations set up as proving grounds for technologies being considered for use on a wider basis. Others are more clandestine. Funded with small amounts of money from other parts of the IT budget, these skunk works operate under the corporate radar and are used, among other purposes, to attract and engage bright and eager young professionals who have little patience with big honking enterprise systems.
"I tend to use the skunk works for things that are way out there," says the CIO of a global consumer products company. "It keeps my technology team engaged."
This CIO, who's the type who likes to keep things under the radar, also notes that a skunk works is an ideal setting to hone IT staffers' understanding of risk. "One of the biggest concepts with risk is the difference between 'I know' and 'I think'. Once you've had the opportunity to play with something, you know about it," he says. The proof is in the projects. "We've increased our project output by 300% because my team now knows how to take risks," he says.
One of his skunk works' biggest successes has been its work with cloud computing. "We're operating an internal cloud and now extending it out to deliver applications to our vendors and trading partners," the CIO says. The ultimate goal is to have all enterprise data reside in the cloud. The IT group will securely deliver applications to any device, but all data will remain in the cloud. Thanks to work done and risks taken in the skunk works, the company is well on its way to achieving this goal, the CIO says.
At Flextronics International, a global contract electronics manufacturer, CIO David Smoley avoided creating a dedicated skunk works in the traditional sense.
"A skunk works is a tool that can help innovation, but they're needed most in an environment that kind of stifles innovation," he says. Smoley recalls his work at previous employers, mainly "large companies where there was a heavy bureaucracy and rigid processes. The way you got around that was to spin off some folks and set them up off-site somewhere."
At Flextronics, he says, "we created a kind of massive skunk works by creating a place where it's safe to take some risks. In fact, we encourage that," he says. "We highlight failing fast, experimenting, trying things but trying them small. If they work, add a little more. If not, throw them out, move on and don't kill the guy who tried it," Smoley says.
This is precisely how the company's internal social network known as Whisper came about. "We have a software group in the Ukraine that developed the tool for their own use. They brought it forward a year ago," Smoley recalls. "There are many companies where I've worked that would immediately have killed the guys because that was not what we were paying them to do. But we embraced it," he says.
Smoley, however, says he wasn't so sure about the need for Whisper when tools like Facebook and Google were already available, so he organised a bakeoff.
"We put together some business and IT folks and ran a pilot and ultimately went with Whisper enterprisewide," he says. "But that never would have happened if the guys in the Ukraine didn't feel safe and comfortable about experimenting. If they felt they were going to get fired, they wouldn't have thought at all about it, and if they did, they would have hidden it."
Yet Smoley is quick to point out that a skunk works or a culture of innovation does not imply or encourage random experimentation, especially during tight economic times.
"We're not just in the business of coming up with cool stuff, but rather coming up with cool stuff to improve security, better customer service and solve other problems," he says. "Innovating with random technologies in an IT department might yield something of business value, but chances are it won't. There's a method to the madness. What I want to do is encourage my guys to innovate around those areas they're responsible for."
Smoley is also a firm believer that proximity to a problem plus a tight economy and constrained resources work to create a climate of increased innovation.
"Some of the most innovative solutions come from IT personnel in the field or factories where they have the dual benefit of being severely constrained and toe-to-toe and eye-to-eye with our end customers," Smoley says. "Those are the guys who have customer service, sales and quality folks in their factories telling them their requirements. As they solve their problems, we troll their websites and look for unique opportunities to spread best practices.
"We also try to take advantage of our lack of resources by encouraging everyone to be innovative," he says. For example, a team in IT built an open source network monitoring tool that Flextronics now uses globally. That tool was developed after officials rejected IT's recommendation that the company buy a multimillion-dollar, commercial network monitoring system.
"The response was, 'We just don't have the money'," Smoley recalls. "But now, we have this open source tool that has literally saved us $10 million."
At JetBlue Airways, CTO Terry Dinterman says skunk works are typically set up to tackle very specific business issues.
For example, the airline set up a skunk works of sorts in direct response to a competitive business need for onboard internet connectivity for passengers.
"Connectivity into the aircraft is a key business driver, and there was a need to get ahead of the industry," Dinterman explains. "So we had a dedicated team from business and technology who were set aside to do some investigation."
Currently, the airline has a small team known as Crew Member Technology Services that is dedicated to "scanning the technology horizon and dabbling with anything out there" in the area of virtualisation and thin client technology, says Dinterman. "By moving to a pure thin client and virtualising our desktops, we expect the cost of supporting customers to be cut by a third," he says. This same group is also investigating the business value of tablets and various "bring your own device" scenarios.
"Pretty much all this group does is try to experiment to make things work and think through the policies and support models and all of the other issues associated with new technologies," he says.
Dinterman says a separate skunk-works-type group is needed, especially when a company is in growth mode like JetBlue.
"We try to be very lean, so asking anyone with an operational, day-to-day initiative to spend a portion of their time dabbling in experimental activities is too much. The pressure of keeping the environment up and running is difficult and unpredictable enough," he says.
Funding a skunk works
For the most part, formal budgets seem to have no place in a skunk works. Instead, most CIOs describe scenarios in which they artfully move money around to fund experiments with technologies that show great promise for delivering business value.
Last year, for example, an IT manager approached Paul Major, managing director of IT at Aspen Skiing Co, with an idea for streamlining an especially labour-intensive and manual process of redeeming vouchers for ski classes with instructors.
"I asked him what he thought he needed and he told me a few hundred dollars for a scanner to see if the idea would work," Major recalls. "The end result is that this year, after a season of experimentation, we're going to roll out a brand-new program that will revolutionise the way we do ski lessons and the way that instructors get scheduled and paid."
As for the budget, he says, "we always have a few line items we could poke a few thousand dollars in without generating too much scrutiny from the finance team. We always will have a way to get would-be experimenters $1,000 and allow them to go play with duct tape and sawdust and try it out without wreaking havoc and see what works," he jokes.
"For us, skunk works is more of a mentality than a process. It doesn't matter if you have $100 or $100,000; it's about giving people resources to try things and see what happens," says Major.
Looking ahead, many CIOs foresee even greater activity in skunk works where they already exist and the creation of new skunk-works-type activities where there are none currently. The reason is directly tied to the rapid pace of technology change.
"There will be much more pressure to do this kind of thing," says the CIO at the consumer products company. "Companies want better, faster, cheaper, but IT doesn't know how to do that unless they try things out."