VentureBeat presents: AI Unleashed - An exclusive executive event for enterprise data leaders. Network and learn with industry peers. Learn More
Joe Kraus isn’t just a Google Ventures partner. He’s also a poker player, and I think he knows that both pursuits involve a certain amount of gambling, regardless of how much skill you think you have.
“In poker, if my analysis doesn’t agree with my gut, I’ll usually go with my gut,” he tells me in his corner office in the Google Ventures building in Mountain View.
“As an angel investor, I made a bunch of one-meeting investments based on feeling. In the venture world, that doesn’t work. … You have a gut read in your first meeting, then you do all of this analysis, and it either confirms your gut or not.”
Related: Read our in-depth interview with Google Ventures partner Kevin Rose.
In our meeting, I wouldn’t have taken Kraus for a gut-feeling risk-chaser. As he speaks, he is incredibly thoughtful and deliberate. He’ll close his eyes, take a breath, introduce a pause in the conversation. I can imagine him in a board room or at a poker table the same way — deliberate, rational, always with a balance in his hand to weigh fact, emotion, ideas, timing, circumstance, and luck. Everything, every word, is measured.
His manner is a good match for the philosophies behind Google Ventures. It has a Googley provenance of big data-meets-brain trust, and its partners operate somewhat differently than other VCs I’ve met in Silicon Valley.
People walk around the venture world thinking they’re such good pickers. It’s like Lake Woebegone, where everyone thinks their children are above average.
Joe Kraus, partner,
Google Ventures is a separate entity from Google, Inc. It operates on the same campus but in different buildings, and while it pulls its talent and knowledge resources from the Google pool, it’s very much its own beast. The fund kicked off in 2009 with a goal of investing $100 million each year. Its known portfolio companies currently number 115; if you look for themes among them, you’ll find they range so widely across any criteria you choose that finding such themes is nearly impossible.
Aside from the Google connection, the firm and its partners are obviously different from anything else in their league in a few major ways.
For one thing, Google Ventures partners don’t really put as much emphasis on the almighty picker: the magic 8-ball in every VC’s back pocket that tells him whether or not a company is a good bet. Primarily because such a device doesn’t exist.
“Picking plays a role, don’t get me wrong. But people walk around the venture world thinking they’re such good pickers,” Kraus says. “It’s like Lake Woebegone, where everyone thinks their children are above average.
“We believe helping companies plays more of a role than most people give it credit for.”
Meet the brain trust
And when Google decides it wants to help a young tech company, the resources it puts behind the effort are mammoth.
Working on artificial intelligence? Google Ventures will likely call one of its advisors, Peter Norvig, an AI expert who used to head up computational sciences at a NASA research center. If your startup is more about energy efficiency, the firm could tap Kenneth Davies to help out; he heads up Google.org’s clean-tech work and the company’s data center energy procurement. Google Ventures also has serial entrepreneur Don Dodge on its list of advisors. It can call up the best of the best hackers, open-source activists, data scientists, you name it.
Part of that is Kraus’s fault. His wiki company, JotSpot, was acquired by Google in 2006, and by 2009 (the year Google Ventures began), he was ready to cash out and go start a new company. In fact, he had a spreadsheet of all the other Googlers he wanted to poach for his next startup.
Instead, Google convinced him to help get Google Ventures off the ground, and his dream-team spreadsheet ended up becoming his first stop for recruiting new Google Ventures team members, people he considered as the best Google had to offer in marketing, design, engineering, and so on.
We’ve all heard those great stories about well-connected angels and VCs making incredible introductions and changing the course of a young company’s fate. But that effect is multiplied many times over for Google Ventures.
“The No. 1 resource is our networks,” says Carrie Farrell, the GV team’s recruitment chief. Farrell spent the better part of a decade fine-tuning Google’s own recruitment process, and in an industry where finding and hiring engineers and designers has practically become a contact sport, she brings some of the most valuable secret sauce of all to the Google Ventures goulash.
“The value-add is immediate,” she said. Looking for a Java developer? “We can give a founder 50 Java resumes in a day.”
“Our biggest problem was recruiting,” said another Google Ventures partner, new kid on the block Kevin Rose. His last company, Milk, had closed a round of funding when it hit a hiring snag.
“We’d had an open spot for an engineer for four or five months, and we just couldn’t fill it,” the Digg founder said. “The candidate flow we got [from Google Ventures], you don’t see that at other VC firms. It was unique enough that I decided to open our round up again.”
That kind of power helps a Google-backed company hire its first five employees and then teach the founders how to hire their next 20. The Google Ventures team keeps a close eye on culture to ensure candidates are a good fit for the growing startup, while prevetting design and engineering candidates with veteran Googlers.
Farrell is just one of the all-stars working on the Google Ventures team. Adam Ghobarah is another. A stats Ph.D with a penchant for big data, his presence saves these startups from having to hire a Ph.D of their own “to tell them where the million-dollar mistakes are.”
After all, he said, “You don’t always need a Ph.D to get value out of the data. But these are complicated things, and it helps to have experience.”
Ghobarah said it’s a “dirty secret in the data community” that the quality of data in these big data sets falls apart due to the way it’s gathered by sensors, server logs, and other machines. “You end up finding patterns that aren’t really there.” But he also says that data quality and insights from the info can prevent startups from flashing in the pan, so he and his big-data cohorts work with 100 percent of the firm’s portfolio companies.
As I learn through spending more time with more Google Ventures folks, this is the firm’s method: Do a lot of hands-on, teach-a-man-to-fish work with founders; dig deep into everyone’s vast networks to get the best help that can be found; and keep pumping time and effort into the portfolio companies, willing them to succeed and generate a return for the firm.
Ron Conway was right all along
After all, this isn’t a summer camp for happy-go-lucky entrepreneurs. The firm is still a venture capital firm, and it has to generate returns. Google being Google, that process started with sorting through a whole lot of data.
Google Ventures employs three full-time statistics Ph.Ds. When the fund was in its earliest stages, the Ph.Ds focused their formidable collective intelligence on the question of early-stage investment. What factors, the partners wanted to know, made for a better chance of seeing a return on their investments?
The statisticians found that the best way to make seed-stage investments and see a profit was to make a huge number of deals — between 60 and 80 per year — that would generate a mean return of 25 percent.
“It was a mathematical validation of Ron [Conway]’s approach from the beginning,” said Kraus, referring to the prominent investor’s shotgun technique of angel investment.
“What really mattered was the diversity,” Kraus continued, underscoring Google’s investment in people over products. “In the seed stage, you’re only gonna focus on the team; the product is too nascent.”
So, when you’re making 60 to 80 investments per year, how do you continue with the hands-on, teach-a-man-to-fish approach? How do you scale that?
Enter Startup Lab, Google Ventures’ take on the incubator
Google Ventures’ newest facility is on the edges of Google’s Mountain View holdings. It’s where the firm is conducting Startup Lab, a program that’s part incubator, part hackathon, part co-working space. It’s how the firm is scaling its hands-on approach to an ever-growing and geographically disperse portfolio of companies.
The Lab’s director is Rick Klau, a GV partner and formerly a project manager for Blogger, Google+, and YouTube. He gives me a tour of the building, which is replete with micro-kitchens and Jenga-like configurations of easily re-arrangeable furniture, strewn about almost haphazardly. Klau tells me that here, you won’t find any paradigms or expectations for how work is supposed to be done. The building will house workshops, talks, office hours, any number of research projects, and even founding teams themselves during the first feverish months of a company’s development.
Klau wants the Lab to create an effect of “osmosis from being surrounded by people who are as good as or better than you at what you’re doing … to meet your peers at other companies and then to talk with someone at Google who’s been doing it for five to 10 years.”
Not that the Lab is all about transferring Google state secrets to a lucky few. Klau emphasizes the non-Googlers that have come to the Lab to give workshops on, for example, Bing web search optimization or the leadership strategies inside Apple.
These kinds of sessions are live-streamed to the firm’s portfolio companies around the globe, and the topics are driven by demand from the founders themselves.
The Lab building contains more than just a live-streaming stage. It features conference rooms with removable, hanging whiteboards that can travel from room to room — a design that was borrowed from Stanford.
The building has a dedicated user experience lab, set up like an interrogation room or focus group room with a one-way mirror, a camera monitoring the room, and separate monitors that show the test subjects’ screens. The UX lab can be reserved for the day, and the studies conducted in it can be recorded.
Startups can make use of a huge boardroom as well, and the office is also a honeycomb of co-working space, set up with 3-foot walls so as not to impede the free flow of natural light and ideas.
Klau relates an anecdote about a portfolio company founder that needed a connection to Pixar, only to find out that a person in the same cubicle had worked at Pixar for four years and was all too happy to make an introduction.
“Cross-pollination … is one benefit of being in the portfolio,” said Klau.
Shoot for the moon, land in a smoking crater?
Still, in spite of the statistics, in spite of the huge investment of resources, venture capital is as much of a gamble at Google as it is at Sequoia or Bessemer or any other firm.
And for Google, that risk is an acceptable one.
“Larry and Sergey are crazy. Larry would rather see a smoking crater in the ground than a mediocre result,” Kraus tells me back in his office of Google’s famous founders, Larry Page and Sergey Brin.
So far, the fund has made fairly early-stage investments — smaller amounts, but with a much greater risk factor. Throwing $1 million at a three-dude “company” that barely has a working product together is cheap (by Google’s standards), but the odds of success are slim, something like 18 percent for first-time entrepreneurs. Throwing $100 million at a proven winner like Instagram is a much surer bet that you, the investor, will get your money back.
Larry would rather see a smoking crater in the ground than a mediocre result.
Joe Kraus, partner,
“It’s really easy to find Series A and Series B companies that are already on a path to success, then you invest in them and double down,” said Rose. “The most exciting thing to me is when it’s just in the wireframe stage. It’s just a seed of an idea, and it’s the boldest thing you can do in investing.”
In those late-stage investments, there’s nothing to build, nothing really to grow. You toss money on an already racing horse that’s several lengths ahead, and you let momentum do its job. A few months or years later, you get a check in the mail. And where’s the fun in that? I don’t think Kraus, Rose, or anyone else on the Google Ventures team would find it to be much fun at all.
“Building keeps me feeling useful,” Kraus said, “like I’m not just — just picking. It’s never boring.”
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.