How Venture Capital’s $215 Billion Is Skewed Toward Software

By Olivia Hart

U.S. venture capitalists invested more than $215 billion across 14,320 deals in 2024, according to a new paper by Duke Law School scholars Emilie Aguirre and Michael Burstein. Software captured 40% of that total, driven by artificial intelligence, while health care and life sciences ranked a distant second at 17.3%.

Aguirre and Burstein report that the venture total exceeded twice federal spending on nonmilitary scientific research and development that year. Their paper, titled Disrupting Venture Innovation, is slated for publication in the Journal of Corporation Law.

What is narrowing the funding mix?

The paper credits venture capital with helping launch Apple, Google, Microsoft and Nvidia. Aguirre and Burstein argue that ideas drive funding, but funding also drives the ideas founders pursue. Aguirre says the market now steers entrepreneurs toward preferred technologies and business models because those promise financial returns, rather than toward the most inventive projects.

Investors chase lucrative exits over short periods, favor entrepreneurs in their own social networks and compete to fund the same technologies and business models at once. As a result, Aguirre says promising ideas may go unfunded, while some entrepreneurs avoid them because they assume venture investors will not support them.

“VC is a real success story that has been great at driving innovation and economic stimulus around the world, yet it is actually funding quite a narrow set of technologies,” Aguirre said. The system is productive, but its concentration can shrink the pool of ideas that reaches the market.

How do funding demands reshape companies?

After eight years of research, Aguirre conducted 150 interviews and observed key personnel at 10 startups across different sectors. She says venture capitalists signal what they will fund through public pitches, social networks and direct contact with entrepreneurs before a financing round; she calls this “inducement.”

Founders respond by changing products, staffing and priorities. Some fired engineers and technical staff and hired more salespeople; others shifted attention toward revenue growth rather than social purpose or product unit economics.

“In sociological or management terms, we would call it isomorphic behavior — adapting yourself to the norm to make yourself look legitimate to a VC funder,” Aguirre explained. She said founders also adapted for the money and the prestige attached to venture backing.

Even funds built around pro-social technologies did not escape the constraints. Aguirre found that impact investors used underlying models fundamentally similar to mainstream venture capital and often reproduced the same behavior. She says companies may forgo genuine innovation and social impact in favor of technology with the potential to dominate a market, or capital-light software instead of hardware, to maximize a fund’s internal rate of return.

The funding process can improve a company’s fit with investor expectations while weakening its fit with its original technical or social purpose.

What could widen the capital base?

Aguirre and Burstein argue that this concentration also damages venture capital’s efficiency. By crowding into familiar bets, funds can miss unique opportunities that might produce outsized returns.

The paper presents several remedies, but not a tested package with published outcome data. Its proposals target fund duration, manager incentives, investor diversity and public participation, while patient-capital funds already operate in biotech.

Aguirre and Burstein propose:

• Establishing more patient-capital funds with longer time horizons, giving founders time to build profitable businesses over an extended period.
• Reworking compensation, which typically equals 2% of fund size plus 20% of profits, to reward portfolio measures beyond spectacular short-term growth.
• Broadening the investor base by addressing barriers to becoming a venture capitalist.
• Exploring government co-investment, as Singapore and Israel do.

The paper names no replacement performance metrics, details no regulatory barriers and assigns no fiscal or risk terms to government co-investment. Wider funding requires longer time horizons, different incentives and more kinds of capital, but the paper leaves the execution largely undefined.

Can corporate law rebalance startup power?

Aguirre says corporate law could also reduce venture capitalists’ power over startups. In 2013, the Delaware Chancery Court held that a company’s directors owe fiduciary duties to the corporation and its common shareholders, not exclusively to venture capitalists as preferred shareholders.

The paper asks whether that legal baseline can make room for the product market, social value and innovation when investors and boards negotiate power. Aguirre also argues that corporate law may need to reallocate power within startups, though she presents this as a question requiring more creative legal design.

Corporate law may restrain funder leverage, but the paper does not identify the rule needed to make that restraint practical.

The next test is whether these levers can expand the set of financed ideas without sacrificing financial discipline. Aguirre and Burstein’s case rests on changing who supplies capital, how long investors wait and what they reward, with law offering a possible limit on the bargaining power of venture funds.