Qualcomm on Building an Internal Startup-Venture Capital Ecosystem

By Stephen Ellison |  March 21, 2016

Many corporate innovation programs struggle to produce tangible impact in areas that really matter to senior leadership. The San Diego-based chipmaker Qualcomm put that challenge front-and-center by dubbing its innovation initiative ImpaQt.

In its first 18 months, says head of innovation Navrina Singh, it delivered a new surround-sound application for HTC smartphones and launched a robotics program to rival Intel, its biggest competitor.

“In 2011, [Qualcomm Executive Chairman Paul Jacobs] asked us to create a new initiative that would be focused on not only technology development but also on activating the change agents within the company so we can bring the right kind of mindset to tech development and problem solving,” says Singh, who reports to Chief Technology Officer Matt Grob. “Basically, that was the genesis of ImpaQt.” The program officially kicked off in 2012.

The ImpaQt blueprint consists of three phases – Ideation, Review and Prototyping – and the process generally takes about nine months. (See below for three slides that offer an overview of the phases.) The program engages the entire employee population, and those who come up with workable ideas remain involved in their projects all the way through to fruition. Singh views it as Qualcomm’s own internal venture capital-startup community. And they embrace many of the elements of lean startup methodology.

“We really operate like an investor … [who] basically mines all these interesting ideas from employees, which are like small startups,” Singh said. “And through funding, through providing a worksite for the teams, we really focus on finding the right product-market fit. Then, once we’ve created the minimum viable product, we scale these ideas either within Qualcomm, or we find the right partners to help us take them to market.”

The biggest improvements ImpaQt has brought, Singh says, is shifting Qualcomm’s innovation approach from building things just for the sake of building to identifying problems, recognizing potential customers, and then putting together the right teams to create compelling solutions. “We have world-class engineers and scientists who can build great things,” she said. “But one of the things that tends to happen in an engineering culture is we focus so much on building stuff that we forget who the customer is.”


Driven by ‘Data and Guts’

The program’s budget fluctuates according to the size and scale of a given project, Singh says, but the starting point is around $1 million. Similarly, the size of Singh’s team varies. She has a staff of five reporting to her directly, and she has an extended “matrix” team that ranges from 100 to 250, she says. The matrix, she explains, is a pool of people that at any given time could include internal business experts; engineers to help build prototypes of the minimum viable product; and external experts from another organization whose product is the focus of an innovation.

“It also varies every year, based on the strategic focus of the program for that year,” Singh says. “In fact, we have a yearly timeline where we come up with three focus areas that Qualcomm actively is looking to create. From those key areas, my extended team expands or shrinks …”

Singh likes to say she is “driven by data and guts.” Each year, her team does extensive analyses, mining the data – the who, what, where and how of ideas brought into the system – to come up with recommendations, changes, and new initiatives. This process, she explains, allows the program to evolve with the latest in market trends and technology changes, as well as the underlying business interests of Qualcomm.

One of the key changes after the first year of the program was narrowing the focus of ideas and their resulting projects. “A lot of people externally say, ‘Let’s think outside the box,’ and we don’t do that,” Singh says. “We actually try to create new boxes and think inside those. What that means is the minute you give creative people a little bit more constraints, they actually come up with solutions that are much more usable. So instead of saying, ‘Come up with an idea in health care,’ we actually give them constraints like, ‘Let’s look at India as a market, where connectivity is limited; you have about $100,000 to operate and two people to work with – what kind of interesting solutions can you come up with for monitoring pregnancies?'”

“The minute you define the problem a little bit more,” she says, “you actually see better solutions and outcomes.”

Creativity at Work

One of those outcomes became a partnership with HTC Corp., whose smartphones, Singh pointed out, are built with a feature not found on many other devices: two front-facing speakers. Using Qualcomm’s patented SnapDragon platform, the ImpaQt team created a technology they call “immersive audio” that provides a surround-sound experience on the HTC One M9 device when users watch YouTube videos, TV shows, or other mobile entertainment.

The idea did not originate with an internal engineer or an external tech partner. It came from a Qualcomm retail product manager who happened to be passionate about quality sound. Robert Dessert, the “idea-ist” behind immersive audio, was given his own team of engineers; they put together a demo and presented it to Qualcomm executives, including Jacobs. So impressed, Jacobs requested the team bring the demo to HTC’s then-CEO Peter Chou, who loved it enough to want it on HTC’s flagship device by the following year. The project went from idea to review to prototyping to fully-functional technology in a span of eight months.

Another area of focus for ImpaQt has been the robotics industry. When the program launched in 2012, there was a lot of work being put into understanding how universities and startups developed robotics projects, Singh recalls. At the time, the dominant robotics operating system, or ROPS, was available only on Intel platforms. Qualcomm sought to get in on the action and, in partnership with the Open Source Robotics Foundation, developed ROPS support on its SnapDragon platform. “So, now,” she says, “if you’re a robotics developer, you essentially have a choice between hardware platforms — Qualcomm or Intel.” Qualcomm quickly has made significant gains against its rival in the robotics space, she says.

“Everyone thinks innovation is this fuzzy, gray, grainy thing, but it’s not.” Singh says. “Innovation is actually making an idea happen and making it a reality…In terms of lessons learned: dream big first, but with big corporations, you have to first show the value – How does it affect the bottom line? How does it affect the employee? How does it affect the shareholder value?”

Singh says that getting the elephant to dance is one of her favorite metaphors for trying to get a large organization to be nimbler and more innovative. “At the end of the day, the minute you make the elephant dance,” Singh says, “what you get is great creativity and great solutions that are actually useful to the firm.”