Breaking Network Barriers in the Era of Data-Driven Venture Capitalists

Venture capital plays a central role in shaping innovation, yet access to venture funding is highly uneven. A small number of regions and industries dominate investment activity, while founders outside these circles often struggle to get noticed. For example in the US, Silicon Valley, New York, and Boston account for the majority of venture deals, and industries such as software and pharma remain the core focus for most investors. These patterns are not accidental. Venture capitalists operate in environments with extreme uncertainty, where startups are young, opaque, and difficult to evaluate. To manage this uncertainty, investors rely heavily on professional networks, geographic proximity, and industry specialization. While these networks facilitate information flow, they also restrict who gets funded.

In a new paper, I examine whether recent advances in data and artificial intelligence are changing how venture capitalists identify investment opportunities. Over the past decade, many VC firms have begun using data platforms and machine learning tools to support early-stage decision-making. Rather than relying on referrals or warm introductions, these tools allow investors to systematically track and evaluate thousands of startups using digital signals such as hiring activity, online presence, website traffic, and founder backgrounds. This shift raises an important question: do data technologies help venture capitalists move beyond traditional networks, and if so, with what consequences?

To study this, I focus on the adoption of data technologies within venture capital firms. I classify a VC as data-driven starting in the year it hires its first data scientist or similar data-focused employee. These hires reflect internal investments in tools that expand deal coverage and standardize early screening. Adoption has risen steadily over the last decade, suggesting that data-driven investing is becoming an increasingly important part of the VC landscape.

The first question I ask is whether data technologies change where and what VCs invest in. I focus on two dimensions that traditionally define VC networks: geography and industry. Venture capital activity in the United States is highly concentrated, with most investments occurring in California, Massachusetts, and New York. Because information is easier to acquire and monitor locally, VCs have historically concentrated investments near their headquarters, reinforcing the dominance of these hubs. Similarly, most VCs specialize in a narrow set of industries, where experience and professional networks provide informational advantages. I therefore examine VCs headquartered in these major hubs and those specializing in software, which accounts for most data-driven VCs in my sample. Startups outside these geographic and industry boundaries are more likely to fall outside a VC’s core network.

The evidence shows that after adopting data technologies, venture capitalists invest more broadly along both dimensions. They increase investments in startups located outside major hubs and in industries beyond their core specialization. On average, data-driven VCs make about 20 percent more out-of-network investments per year. Importantly, this shift is not driven by an increase in overall investment activity. Instead, portfolios tilt toward startups that would previously have been less likely to receive funding. These effects are strongest where traditional networks are weakest. Data-driven investors are particularly more likely to fund startups in regions with little prior VC activity, founders who did not attend elite universities, and first-time entrepreneurs. 

A natural concern is whether some other force drives both data adoption and out-of-network investing. One possibility is overall firm expansion: growing VC firms may both hire data specialists and invest more broadly. To address this, I compare data-driven VCs to firms that hire a new general partner in the same year. Both groups are thus expanding. I find that data-driven firms are no more likely to increase their total investments than this comparison group, yet they are significantly more likely to invest outside major hubs and core industries. To further rule out unobserved differences, I exploit the interaction between VCs’ prior exposure to data technologies and their current capacity to hire, proxied by recent fundraising. The results remain unchanged.

Expanding the investment net raises an obvious question: how do these out-of-network investments perform? Investing far from established networks may reduce an investor’s ability to monitor startups, provide advice, or intervene when problems arise. At the same time, lower competition outside core networks may allow VCs to access higher-quality opportunities.

To answer this, I examine how data-driven investments perform relative to traditional investments. On average, data-driven investments are more likely to receive follow-on funding, suggesting improved early-stage screening. However, this advantage does not fully extend to out-of-network investments. Startups funded outside a VC’s core geography or industry are no more likely to receive early follow-on financing than in-network investments, consistent with monitoring being more difficult in these settings.

The picture changes when focusing on long-run outcomes. Conditional on surviving early funding rounds, out-of-network investments made by data-driven VCs are more likely to achieve major successes, such as IPOs or high-quality acquisitions. These investments outperform both in-network data-driven deals and out-of-network investments made by traditional VCs. This pattern suggests that while monitoring challenges remain, data technologies help identify high-potential startups in less crowded markets. When these startups succeed, they do so at disproportionately high rates.

The final part of the paper studies the real effects of data-driven investing. When a data-driven VC makes its first investment in a region with historically low VC activity, that investment can act as a signal. It raises the visibility of local startups and connects the region to broader investor networks. Following such entry, these regions experience meaningful increases in venture activity, including more startups receiving first-time VC funding, more distinct investors, and greater innovation output. Across measures, subsequent VC activity rises by roughly 10 to 30 percent.

Taken together, these findings show how data technologies are reshaping the venture capital landscape. By reducing reliance on closed networks, data-driven tools expand who gets seen, who gets funded, and where innovation takes place. They do not eliminate all frictions as monitoring remains more challenging for startups further from a VC’s core geography or industry. When those challenges are ultimately overcome, however, these out-of-network investments are high quality and disproportionately likely to achieve major exits. Overall, this paper shows that data technologies can meaningfully expand the reach of venture capital.

-----------------------------------

Melissa Crumling is a Ph.D. Candidate in Finance at Drexel University's Lebow College of Business.

This blog is on the paper presented at the 6th Annual RCF-ECGI Corporate Finance and Governance Conference, held in a hybrid format online and in Hoboken, New Jersey, on 13 and 14 December 2025. Visit the event page to explore more conference-related blogs.

The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.