Data Science in Venture Capital
In the past few days there has been a bit of a meme emerging on the topic of applying data science to venture capital — including an article in VentureBeat, and a few posts about Balderton Capital’s new data scientist role.
Some of the coverage has mentioned that we are hiring for a data scientist role at General Catalyst (or as we have taken to calling it, a “data associate”). So, I thought I might briefly chime in with my own view on the role of data science in VC.
Did everything just change?
I see data science in VC not as a fundamental sea change, but rather as an exciting way to augment the traditional venture investing process.
While some interesting new firms are taking very aggressively data-centric approaches, others are applying data science techniques to support an existing core approach. At General Catalyst, our primary investment thesis remains first and foremost people centric: we are focused on partnering with forward-looking and inspiring visionaries who can lead the next generation of industry-changing companies. In our worldview, data science potentially provides a means for achieving that end.
A theory on where data science fits
My friend Rob Go at NextView Ventures has previously written about the venture process in terms of 3 stages: source, select, and win.
Borrowing Rob’s framework, data science has potential applications to each of these:
Source: This use case seems to be getting a lot of attention. With today’s high volume of startup activity, it’s not always easy to find and prioritize the relevant projects that match a particular firm’s strategy. Data-driven approaches can provide additional coverage, but probably even more importantly, help to prioritize the investment pipeline and focus attention on the highest potential opportunities. Here there are intriguing analogies to CRM pipeline predictive scoring products like Infer.
Select: Venture Capitalists have always conducted quantitative due diligence on investment opportunities, typically focused on financials and key business metrics. Data science can be applied here to provide additional, rigorously supported quantitative insights in areas like growth analysis and competitive benchmarking.
Win: Data science talent is in high demand and costly for early stage startups to acquire. By making data science capabilities and insights available to portfolio companies, venture firms have an opportunity to distinguish themselves from the pack in terms of value add. Beyond data scientist consultations, one can imagine the creation of durable data products that are shared across a firm's portfolio companies or a wider community. To the extent these capabilities deliver novel value, they will help firms win deals, and more importantly, help portfolio companies win in the market.
Some things never change
At the end of the day, I believe the most important element of early stage ventures is still the audacious and magnetic entrepreneurs who make it happen against all odds. I don't expect to see data science upset this fundamental truth, but it certainly has the potential to help inform the investing process.
Data science in venture capital is still in its early days. It will be exciting to see how today’s early experiments pan out.
If you are potentially interested in being part of our experiments with data science at General Catalyst in Palo Alto or Cambridge, you can learn more about the role via LinkedIn.