He may be one of the brains behind a multi-billion-dollar business, but Google Cloud CTO Will Grannis hasn’t always been in a position of privilege and power.
Growing up as the middle of three children in a single-parent household in Phoenix, Arizona, Grannis said he learned what government-issued powdered milk tasted like. He experienced what it’s like for a family to lose all their possessions in multiple bankruptcies. That upbringing left him not only with a desire to smooth the path for others, but also the resilience and determination to get back up after multiple failures and try again.
What does that have to do with Google Cloud? A lot, actually. Failure, Grannis told Silverlinings in an exclusive interview, plays a big role in Google Cloud’s Office of the CTO (OCTO).
Check out a clip from our chat with Will Grannis in our Lunch Ladies podcast here!
Formed in 2015, OCTO comprises a team of senior Google technology executives from 20-plus countries who think big about where Google Cloud should invest next. They work with the top 150 companies using Google Cloud to test trends and big ideas as they bubble up. But Grannis noted their ideas don’t always pan out. For example, if OCTO chooses to chase 10 big ideas in a given year, maybe three will actually be viable, he said.
“Successes are visible. The failures are more numerous,” Grannis explained. “We have to go into it with a mindset of exploration, testing and failing. We fail all the time.”
But it’s a balancing act between knowing when to let failures go and knowing when a truly innovative idea just needs time to develop. That’s not always easy given businesses – including Google Cloud’s – operate with a need to deliver short-term results.
“Some of these things take years to realize. AI is a really good example,” he said. “Neural networks as a concept had been around for you could argue decades. The reason why in 2015, 2016 they really rose to prominence is because computation became feasible from a performance and cost trade off to actually run them at scale. So, it’s hard to predict when those shifts are going to happen.”
Grannis added, “One of the things that I’ve learned over time is that when innovative ideas are truly innovative, nobody is talking about them.”
He pointed to sustainability as an example. Today, sustainability is a critical initiative for every business, particularly cloud companies. But five-plus years ago, it wasn’t necessarily at the top of anyone’s priority list, he said.
AI revolution
It’s no secret that artificial intelligence (AI) is at the top of Google Cloud’s list of priorities today. Grannis is tackling this tech as he juggles his own personal priorities — his wife and two daughters, one a high school senior and the other in college.
A self-proclaimed "soccer dad," Grannis said he's grappling with the fact that he's "on the cliff of having my life’s purpose walk out the door" and what the future holds for him as an empty-nester.
In his professional life, Grannis is busy wrapping his arms around what the future holds from a technology perspective. Grannis painted a picture of a world where AI becomes “part of almost every service you can imagine” over the coming years. One in which AI moves from a novelty to a technology that transforms the lives of citizens each day via improvements in government, finance, commerce and healthcare. One in which there are continuous improvements in sustainability such that AI can be delivered using less power.
Some of these ideas are objectively cool to contemplate. For instance, he sketched out one AI use case in which the technology is used to provide auto-charting and note-taking for doctors to give them more time with patients.
On the sustainability side, he said the concept of “least-cost routing” could help right size cloud resources to use less power by automatically matching the needs of an AI workload to the right model.
But like all good things, Grannis said this revolution will take time. Likewise, it will also take time for the investments Google Cloud has made in things like AI safety, security and responsibility to be recognized as critical in nature.
Grannis said he already sees the industry starting to exit the AI hype cycle – that is, companies have now experimented with AI, know how they want to use it and are taking pragmatic steps in that direction.
“So, with that passage of time I think a lot of places where we’ve been investing and market leaders will actually rise in importance,” he said.
Getting to Google
OCTO’s influence on the direction Google Cloud takes is still fairly new and Grannis’ CTO post is an even more recent addition. For the first five years of existence, OCTO functioned without a formal CTO, outsourcing the role instead to the experts who made up the office. Grannis was named CTO of Google Cloud in 2020 and, by his own admission, he wasn’t the obvious choice for the role.
“I am not from tech. I don’t come from the tech industry. I come from enterprise, I come from industry, I come from applying technology,” he said. “It was a non-intuitive choice to pick someone from outside of that usual circle.”
Grannis may not be your typical tech-bro type, but he’s always been fascinated by technology and how things work. He recalled that while gaming as a kid, he was captivated by the idea that a spinning disc in a console could render an entire world responsive to outside inputs. He wondered how the signal was sent from the controller to the TV, what the processing mechanism was and just how it all worked.
That curiosity, coupled with a desire to serve others and to build things, has defined Grannis’ career path.
He got his start in the military, serving five years in the Army. There he found technology was actually “a barrier and a risk to the people who signed up” and determined that the best way to effect change was from the outside. So, he moved to a small company building satellites, did a six-year stint with Boeing Research and Technology and eventually started his own company with the goal of – in all seriousness – using data to pinpoint the supply chain for weapons of mass destruction.
After an acquisition, he became the CTO of defense company L3 Technologies, where he remained until joining Google in 2015. It was in his role at L3 that he bumped into the technology that convinced him he “had to get to Google”: BigQuery.
For those not in the know, BigQuery is Google’s data warehouse service that lets users run analyses over petabytes worth of information.
Discovering the ability to do that for “pennies” was a game-changer for Grannis. It was BigQuery that convinced him that Google was “building the future of analytics.”
Multi-cloud is underrated
During Grannis’ tenure at Google Cloud, BigQuery has become part of another major – and, according to Grannis, underrated – initiative at the company: multi-cloud.
“One of the things I really am impressed by and still trying to figure out how to leverage its full utility is a concept that we’ve been building here for a while at Google Cloud which is multi-cloud,” he said. “That may sound like a buzzword but we’ve done a lot of engineering to make it so that customers can use the cloud and use capabilities from our cloud across data stored in any cloud.”
He pointed to the launch of its multi-cloud orchestrator, Anthos, in 2019, as well as the subsequent general availability debut of BigQuery Omni in October 2021 and AlloyDB Omni earlier this year.
“There’s a consistent thread running through all of them and that is that we really, really believe that data, databases, analytics is the lifeblood of a digital organization,” he concluded. “As a primary cloud provider, we’re the only ones providing true multi-cloud databases, true multi-cloud analytics, true multi-cloud orchestration. So, now you can imagine where we’re headed next is around AI.”
Ed. Note: Grannis named a number of individuals who have mentored him, influencing both his career path and leadership style. These include General Brad Becker, who taught him to lead from the front; Comtech's Tim Moore, who taught him about the business of making technology viable; Bob Logan; Boeing's Dennis Muilenburg; Waldo Carmona, who helped him learn to focus on capabilities that actually make a difference; Roger Krone; former CIA CIO Alan Wade; and various Google staff including Carl, Brian, Urs, Diane and Thomas, all of whom apparently lack last names.
Read more inspiring profiles of tech leaders here.