Today at Web 2.0, John Batelle sat down with Bill Maris and Graham Spencer of Google Ventures for a quick peek inside Google’s venture arm. For those unfamiliar, Google Ventures’ team of 43 invests up to $200 million a year and in 2011 will make upwards of 100 investments. The value proposition of Google’s venture wing is that, according to Maris, it’s highly focused on engagement with the startups they’re investing in, and, of course, partnerships and Google resources are always close at hand. While Google Ventures doesn’t invest specifically to further Google Inc’s products or development projects, the team consists primarily of former Googlers — not to mention former entrepreneurs, investors, and scientists.
But what’s probably most interesting about the venture firm is that it’s really taking a Moneyball approach to investing in startups. What does this mean? Moneyball, the book (and now film) that looks at the changing role of statistics in major league baseball and how clubhouses utilize their data to determine the value of players. Billy Beane, the book’s protagonist, was able to build an Oakland Athletics team of undervalued (and many would say questionable) players for cheap and take that team all the way to the playoffs.
Google Ventures, too, is all about the data. Beyond being able to help teams solve their UX, search, and scaling issues, the team takes hopes to mix both qualitative and quantitative data to put together the best approach to deciding in which startups to invest. According to Maris and Spencer, Google Ventures applies data tools and metrics to their prospective investments to decide which opportunities make the most sense.
Naturally, as the team at Google Ventures itself (and their mothership as well) is home a full roster of computer scientists, engineers, machine learning specialists, and more, there’s really no surprise that Google Ventures would be applying a data frame to the problem of what spaces and areas are most ripe for investment.
They take investment heuristics and the many qualitative rules of thumb investors use to make their decisions, and apply software they’ve built themselves to build models, test theories, etc. So along with being able to take advantage of Google-based analytics, search query-type knowledge, and so on.
The venture partners cited DNAnexus as one of the companies they’ve invested in most recently that both reflects their data-based approach as well as their strategic effort to find startups that are designing solutions to big, global problems. DNAnexus is building a backend system to help researchers understand genomic data sets, making that reseaerch available to everybody, not just for Google — even though Google Inc has partnered with the company to host this huge, totally anonymized, government database.
The startup is a good example, the venture partners said, of the type of startup they’re looking to invest in, whereupon they listed a few more examples, including Climate Corp (weather insurance for farmers), and Adimab, an antibody discovery platform.
For more on Google Ventures’ portfolio, check ‘em out here.