SignalFire: How Machine Learning and Data Science Are Changing VC
--
It’s strange how resistant people are to change. What’s even weirder is that sometimes, the people who are most resistant to change are those who champion it most strongly in others.
Picture the traditional VC firm of the 1960s-early 2000s. Deal flow would primarily come through your network, or if you were lucky enough to work at a sexy brand like Sequoia or KPCB, you could see a lot of inbound deals. Thus, the two things venture capitalists competed on were brand and relationships. The only innovation to the industry during this time were increasing fund sizes.
Then something interesting happened after the dot-com crash. A host of firms that were heavily focused on “value-add” appeared, most notably Andreessen Horowitz. In addition to providing capital, these funds also helped entrepreneurs with strategy, hiring engineers, raising additional rounds of funding, and more. And as the funds of old grew larger and larger, a host of earlier-stage funds like First Round appeared. During this period, venture capital firms expanded in their breadth of services and many more funds at all stages of the fundraising cycle (seed, Series A, B, and so on) joined the party. While brand and relationships remained paramount, how much value the firm’s partners could add to a company began mattering a lot more.
Today, there are so many VC firms that it is becoming increasingly difficult to tell them apart. Funds like Forerunner have done very well in specializing in a niche sector, and other firms have differentiated themselves by building pools of people (scout programs, accelerators, student venture ambassadors, etc.). to help identify hot startups.
However, there is only so much one can do to increase deal flow or value-add when you rely solely on people. Andreessen Horowitz employs 100+ people to help portfolio companies with operations and they’ve done extremely well, but what if you could do the same thing with one person behind a computer instead of a team of recruiters? Or what if instead of looking at 5,000 startups per year, you could increase that amount to 100,000?
Enter SignalFire.
Founded in 2013 by Chris Farmer (ex-General Catalyst partner) and Ilya Kirnos (ex-Google), SignalFire is a data-centric fund that leverages the wonders of software to identify promising deals and support portfolio companies. To date, they have raised $330 million across two funds to power their unique investment approach of combining data science with human intelligence.
Armed with technology, SignalFire can filter through deals with greater pace and accuracy than traditional VC firms. At any given moment there exists thousands of startups that could intrigue a VC, and by following markers like app traction, whether key employees have left, founders’ previous success(es), etc. for hundreds of thousands of companies, SignalFire is able to quickly filter that number down to a much smaller figure. The firm’s investors then take an intimate look at the promising ones. Andrew Ng, who joined SignalFire in 2018, states that, “While computers are far better than us at narrowing the scope of what we should look at, we’ve found that humans are much better at actually evaluating the 10–12 promising companies identified by the computer.” Whereas many firms meet with hundreds of companies (most of which end up going nowhere), SignalFire is able to take far fewer total meetings with an equal if not greater number of fruitful ones. That is the power of technology, applied to venture capital.
In addition to sourcing, SignalFire also uses software to enhance the VC service-provider model pioneered by Andreessen. They have built proprietary platforms that enable portfolio companies to scale engineering, manage talent, etc. The firm’s team reflects this dedication to data-informed investing and value-add. Andrew, for example, spends ⅔ of his time on the investment team and the remaining ⅓ in engineering.
So, what does all this mean for the future of venture capital? Is the industry one day just going to be a bunch of programmers sitting behind a computer screen, eating chips and watching numbers?
The answer is: of course not, such a scenario is ludicrous. Data and AI will never replace humans in the world of venture; they are merely a means to an end. Founding partner and CTO of SignalFire, Ilya Kirnos, described AI as standing for “augmented intelligence,” meaning data is simply a tool that gives humans superpowers.
Real recognizes real, and it is important to recognize that the human element will always remain a part of venture. A founder’s passion, drive to succeed, charisma… these cannot be quantified. In response to a question regarding the firm’s attitude towards gut judgement of a founder’s intangibles, Ilya Kirnos responded:
“There are many qualities of a person that you can only discern from sitting across from them. The things we look for in meetings are whether founders are magnetic, able to recruit, and sell. Data gives us a huge edge, but at the end of the day we are human investors that close the deal.”
The data-driven model has shown promise for SignalFire and a handful of other firms, and slowly the industry is moving in that direction. However, significant roadblocks remain. It is a serious investment for firms, especially incumbent ones, to hire a team of engineers and data scientists to implement this radical new approach. When Ilya and Chris were still exploring the idea, many VCs told Ilya that SignalFire’s strategy wouldn’t work. You can imagine that a successful GP might think, “My way has worked well in the past, so there’s no reason to change.”
However, that is precisely the same mentality that allowed so many venture-backed startups to destroy comfortable incumbents over the years. The whole business of VC is to be optimistic about innovation, to upend the status quo. VCs often question how scalable a startup’s business model is or how software can revolutionize an industry. It is a little ironic then, that by and large they have not looked at how software can improve their own business.
I will leave you with a scene from The Office. When asked, “Sir, as a company that primarily distributes paper, how have you adapted your business model to function in an increasingly paperless world?”, Michael Scott responded:
“We can’t overestimate the value of computers. Yes, they are great for playing games and forwarding funny emails. But real business is done on paper, okay? Write that down.”
There’s always room for improvement.
—
Special thanks to Andrew Ng and Ilya Kirnos from SignalFire (and Princeton Class of ’18 and ’99, respectively) for their contributions to this piece.
Hey, are you a college student who’s interested in writing about entrepreneurship and talking to founders/VCs? Then you should join the PIE team!
