Bridging Corporate Venture and Independent VC: Eric’s Journey, Investment Insights, and the Future of AI & Startups

1. Eric, you have an extensive background in finance and tech, having worked for major corporates like Oracle and C3 AI. What inspired your transition from corporate finance to venture capital?

I had the privilege of working at Oracle with legendary venture investor and board member Don Lucas, who had provided the seed capital to Oracle in 1980 to get it started, and went on to invest early in many other winners. I witnessed directly the outsized effect that a small amount of capital and a bit of time could have on the growth trajectory of a young company.

Then at Oracle we raised $52 billion in debt capital and acquired 103 companies during my tenure, and it turned out many of them had corporate venture arms that had invested in startups themselves. We inherited a corporate venture portfolio that needed to be managed. I had the opportunity as a corporate VC to participate in the Kauffman Fellows program for training venture capitalists. There I met my current business partner Jack Crawford, and we saw an opportunity to provide a link between corporate venture arms and independent venture capital firms. And I also saw an opportunity to step away from a supporting role of a highly-successful corporate team and into a role making all of the key decisions in a smaller organization.

2. What was the main gap in the VC landscape that led you to co-found Impact Venture Capital?

I had some formative experiences in corporate venture.

In 2007, I approached Larry Ellison and suggested we launch a corporate venture arm. He initially was cool on the idea, noting that major VCs often treated corporates as “the dumb money”. After we acquired 103 companies from 2005-10, it turned out that many of them had their own venture arms, and by 2010 we had investment stakes in over 180 different startups. I shared this with Larry, and his response was “well if we have an accidental corporate venture arm, let’s maximize the value” and empowered me to make follow-on investments. The portfolio ended up generating value. But I noticed that when I approached the largest VCs on Sand Hill Road, they were usually enthused to take the calls and pitch their portfolio companies for Oracle to acquire. But they were less receptive to coinvestment in their young startups. I perceived that they thought a corporate making startup investments was not ‘staying in their lane’.

This is the origin story of what motivated me to start a venture firm after 30 years in corporates. I wanted to launch the corporate-friendly independent VC. At the time, the percentage of venture-backed deals with a corporate participant was quite small, less than one-tenth of venture deals had a corporate investor. That has changed dramatically. In 2024 one-fourth of deals had a corporate investor. And the percent-of-dollars invested was well over half (as corporates emphasize quality over quantity).

This is a key trend in venture that does not get enough attention. Corporates no longer rely solely on internal R&D and acquisition to get access to new technology, but now rely also on the ‘third leg’ of the stool which is coinvesting in young startups to stay abreast of innovation faster.

Venture Capital as an Asset Class

3. You emphasize the importance of venture capital in family office portfolios. Why do you believe venture should be a key part of their investment strategy?

Venture always has attractive features in an individual or family office portfolio.

•  Venture has a wider dispersion in returns than other assets, but over long periods of time has the highest return of any asset class. This short-term variance means most of us should only invest a  fraction of our wealth not needed for liquidity
• However, venture has an imperfect correlation with other assets, so can lower the variance of portfolio return even with its individual variation
•  Within venture, early-stage has higher returns than later-stage venture

So venture has the dual characteristics of a higher long-run average return and an ability to offer a volatility-dampening diversification component to a family office portfolio across multiple asset classes. Over long periods of time, venture has been a great component to a family office portfolio.

4. Over the past three years, capital deployment into venture has declined. What factors have contributed to this, and why do you see this as an ideal time for new investments?

The amount of dollars invested by US venture firms into startups peaked at $350 billion in 2021. This was driven by a huge influx of capital into ever-larger venture funds from US institutional LP investors (endowments, fund-of-funds) who saw other institutions with high returns in the asset class. As capital ballooned, so did valuations. The percentage of portfolios comprised by the venture asset class increased.

However, after this runup, many US institutional investors felt overallocated to the asset class and then started pulling back, and the venture firms themselves started stockpiling dry powder. Both in the US and globally, the amount invested by venture firms into startup companies fell by a third in 2022. Then, the amount invested into startups fell by another third in 2023. So in just two years, the amount of capital deployed by VC firms fell in half. This has not changed in the past five quarters since the end of 2023.

Incidentally, VCs continued to raise their own capital for another year after they slowed deployment, so they were still raising through 2022. By 2023 however, fundraising followed suit, falling 60% in 2023 alone and another 8% or so in 2024. So VC fundraising is down two-thirds in the past three years.

In public equity markets, I have seen enough research to believe that few can time that market consistently. But in private markets, there are more ebbs and flows in the aggregate valuations. Since the end of 2022, so little capital is flowing to startups that valuations are significantly down. This is creating an environment that PitchBook last year called “the most investor-friendly environment of the last decade.” This means that investors in startups have the power to set the terms, including investor-friendly valuations as an entry point for investment. Which we conclude makes it a good time to be a new investor in startups and the funds that invest in them. It’s a particularly good entry point for those players that weren’t already heavily allocated to the asset class, like family offices considering entering the asset class today.

There is another reason why now may be an opportune entry point. We looked at some historical data about how venture fared over a number of years, and the data show that venture has been a better investment in years soon after a macro disruption. This was true after the crises of 1987, 2001, and 2008. We think it is true today after the recent downturn in the general venture environment (which has been partially offset by the excitement around AI business models). Crises lower valuations, crises passed and valuations rebounded, and investors from that period get additional return.

For example, 2008 saw the seed rounds at low valuations for several huge eventual winners:  Stripe, Square, Uber, Lyft, Airbnb, GroupOn.

5. What are some misconceptions family offices have about venture investing, and how do you address them?

I think some family offices have concerns about venture (particularly early-stage venture):

• (1) ‘It is too risky’: Venture has a higher standalone variance than other asset classes. This is true, and represents a good reason not to put all your wealth in to this single asset class. But that misses the benefit of diversification discussed earlier. Because private stocks do not move in concert with public stocks, the imperfect correlation means that the individual variance does not necessarily translate into a higher portfolio variance.
• (2) ‘It is not liquid’: This is true, investing in a venture fund can tie your commitment up for up to a decade. Another reason not to put all your capital in a single asset class. But the historical returns are so much higher than other asset classes that it pays well if your investment horizon allows you to have a modest percentage of your wealth illiquid. Do you really expect to need 100% of your wealth converted to cash in the next decade? 
• (3) ‘It require special expertise’: Conceptually, it is just another category of stock in companies, just stock that does not reveal a trading price daily. And it’s a larger pool of candidates. There are 3450 public stocks on Nasdaq and 1840 public stocks on Euronext. There are many times as many private stocks (among just US companies with $100 million in revenues, public stocks are just 13% of the total). And if you invest in venture funds, you can outsource that expertise of evaluating private companies. Investing in funds is a good way to gain experience, and deal flow, to assist family offices that want to start investing directly into startup companies.

The Role of AI in Startups

6-7. AI is reshaping industries across the board. From your experience, how has AI evolved in the startup space, and where do you see the biggest opportunities for innovation? What sectors will be the most disrupted by AI in the next 5-10 years.

AI is becoming THE dominant area for venture investment. When we started in 2016, less than one-tenth of venture deals were AI-oriented. In 2024, over a third of the companies by deal count, and over half by dollars invested, had a business model based on artificial intelligence.

AI is an overly broad term. With many startups claiming to be “AI” to harness the trendiness, it is important to distinguish between specific categories.

ChatGPT was unveiled just two years ago at the end of 2022. It showed the power of large language models, and OpenAI became extraordinarily valuable. Then in the last year DeepSeek showed that many of these same capabilities could be achieved at a far lower cost. Large language models appear to be becoming commoditized, and the big money has arguably already been made there.

We see potential for new use cases of AI. Some of these we put in the ‘picks and shovels’ category of just providing foundational ingredients like compute power and speed (as with our hardware portfolio company Cornami using faster chips to create better cybersecurity). In other instances, it is taking an existing AI or machine learning tool and applying it in a new sector or industry (as with the auction platform developed by our portfolio company  CapConnect+ to make corporate debt issuance more efficient). To stretch the gold mining metaphor, these are the ‘jewelers’ and an area of focus for us.

I would also note that the growth in AI-oriented tech startups, and in the participation by big tech investing in startups, are actually not separate phenomena. They are tied together. The advent of more AI related startups is pushing corporates to invest more, and is a big part of the $108 billion in US venture deal activity in 2024 that involved at least one corporate investor.

8. How does Impact Venture Capital identify promising AI startups? What criteria do you prioritize when making investment decisions?

The National Venture Capital Association estimates the number of venture firms in the US to be about 3400, which includes all types. If we exclude corporates and one-person shops and niche specialists, the number is noticeably smaller. It’s also smaller in Europe. So if you are the founder of an early-stage AI-oriented startup, it’s not very hard to identify most of the VCs whose investment thesis might align with your business model. At Impact VC, we are getting a couple of thousand inbound inquiries each year. So it’s not finding the startups that is a challenge as much as filtering for the good ones. At our firm, we look first at those startups referred to us by our own limited partner investors, and second to those referred to us by our corporate partners, and third those referred by our broad network of advisors and coinvestors at other firms.

Like most VCs, we look at: the team, the technology, the barriers to being copied (whether from intellectual property or first-mover advantage), and the addressable market. Venture investing is built on outliers, which means most VCs will see a disproportionate percentage of a fund’s overall return come from its single most successful company. This is the ‘power law’ described well by Peter Thiel in his excellent book “Zero to One” (a great primer on the thought process in venture). So it’s important not just to look at the odds of success, but rather how good can the best-case scenario get? The great thing about venture is that you can earn 100 times your investment on a good company, but won’t (normally) lose more than 1x your investment on a bad company.

We focus a lot on team. Do they understand the risks and opportunities of being a startup? Do they have a mix of skills useful for a new startup, and also do they have anyone there who knows how to operate in a bigger company so that they can handle initial success? Importantly, how coachable is the CEO? It can be hard to find the right combination of confidence and humility required to grow a company from nothing to exit-ready. Do we have the right personal chemistry to be confident that we can have a constructive discussion under stress? These are some of the questions that we address in our internal investment committee discussion around investing in an individual startup.

We also add a couple of elements that may be more unique to our own thesis. We ask is this truly using artificial intelligence or is this software with a new label? Are there other use cases for the tools involved that can expand the addressable market? Is this a company and technology that would be interesting to our corporate partners?

Corporate Collaboration in Venture

9. You’ve mentioned that many VCs focus on selling startups to big tech rather than collaborating early on. How does Impact Venture Capital approach corporate partnerships differently?

Most, over 90%, of successful exits take the form of selling to a larger acquirer. IPOs do happen but they are less common even for successful outcomes. So almost all venture investors hope that they can get interest from corporates, hopefully from more than one so that the price gets bid high to create nice returns for the investors. Historically, many corporates are active acquirers of startups as a way to bring new technology into the company.

Before ten years ago, relatively few corporates made venture investments. That is, they might buy 100% of a startup but less often would buy 20% of that startup. These days that is less true, though some corporates still have a minimum investment level that keeps them less active in investing at the seed-stage.

At Impact VC, we talk to corporates regularly, even before we have a specific company in mind for coinvestment. We try to understand their tech priorities for that year, what types of companies might they be interested in and what types are they less likely to be interested in. So that when we evaluate a potential seed-stage startup, we already have some idea which corporates might eventually become interested when that startup matures a bit. It sometimes goes the other way, we have corporates approach us and tell us about a company they like that is ‘too early’ for the corporate, and the corporate asks if we want to invest early, then help the startup mature, then bring it back to the corporate for a later-stage round.

Because of this partnership, with the investment of less than $50 million in our first two funds, we have catalyzed follow-on investment from corporate partners (and larger VCs) of over $500 million. We consider the venture lifecycle to be a relay race. The first leg might be the founder’s own capital and friends-and-family (preseed), we might lead the second leg of the race as the first institutional investor, and then we recruit a corporate to invest in the third leg in an ‘A’ or ‘B’ round.

10. What are the key advantages of corporate-startup collaboration, both for the startups and for the corporates?

For us, the key advantage is an informational advantage. If the corporate venture arm is already invested at the ‘A’ round or ‘B’ round, and is already familiar with the startup, we think it’s more likely that a corporate will make an acquisition bid for the company later. It is a way of mitigating the risk of investing for us. We don’t require a corporate to be an LP investor in our fund (though some choose to), and many tech corporates don’t act as fund investors. But we see the corporates as an ingredient to the eventual successful exit of the company to benefit all our investors.

The increased likelihood of exit helps the startup as well as ourselves. But the startup also benefits from understanding how the corporate, as both a customer and potential later owner, thinks about the technology. And often the corporate becomes an anchor early customer for the company when it needs to generate revenue, and provides assurance to other customers that the company will stay alive to provide its products or services.

11. Can you share an example of a successful investment where corporate collaboration played a critical role?

Corporate collaboration has been central to the growth of several of our portfolio companies. We invested in 2016 in an early round (at a low valuation) of the hardware startup Cornami that was developing a data-center-on-a-chip, that could retrieve and manipulate data far faster than existing technologies. We recruited Baidu Ventures to co-lead the larger next round. And then this year SoftBank led a new Series D round at a much higher million valuation.

We were also early investors in a company Airlinq that provides a mobile internet-of-things platform that allows cars to communicate with their manufacturers in real-time as they go down the road. General Motors and Verizon Ventures both invested early, and Radiomovil invested later, providing necessary capital to the growth of this promising telecommunications startup.

For us, this dynamic is the norm and not the exception. Which is the foundation of how we turned $50 million of our own investment into over $500 million of follow-on capital for our portfolio companies.

SPACs & Public Markets

12. You are chairing the Archimedes II Tech SPAC, and previously, you took SoundHound public via SPAC in 2022. Why do you see SPACs as a viable alternative to traditional IPOs?

Startups seeking to go public can go the traditional path, working with an investment bank to prepare an IPO, building the finance team for the rigors of public reporting requirements, and working to satisfy the intense initial reporting and disclosure requirements around historical and projected financials mandated by regulatory authorities (like the SEC in the US). It is a long and complicated process.

A special purpose acquisition corporation (SPAC) simplifies this process somewhat. The SPAC goes public while still simply a shell with no assets other than cash, which simplifies the disclosure needed for the IPO process. Then this public shell company partners with an operating company to merge (which has disclosure requirements that are involved but not quite as stringent as with a traditional IPO), with the partner company taking on the publicly-traded combination.

• These offers a few benefits versus a traditional IPO. First, it can be done faster. In fact, the target partner controls any conditions that are set to execute the combination, which reduces the uncertainty about whether going public will actually happen.
• Second, it avoids a ‘failed IPO’. The target company controls the timing, and avoids the risk and embarrassment of an IPO that does not happen.
• Third, it can work even if markets do not cooperate. The traditional IPO process tends to go through ‘open’ and ‘closed’ periods when no, or few, companies are able to complete a traditional transaction. With a SPAC, it is technically a merger transaction and less prone to these ebbs and flows.
• Fourth, the process offers more flexibility than a traditional IPO. The partner company can take more time with a SPAC merger process to raise the company’s profile, educate the market, attract analyst coverage, and find the right investors.

In 2022, the Archimedes Tech SPAC team partnered with conversational AI unicorn SoundHound to take them public as the surviving entity in a merger with the already-public SPAC entity. This worked out well, and the market cap of SoundHound was $3.8 billion on March 24th.

13. What types of startups should consider the SPAC route, and what should they be aware of before going public this way?

Startup management teams that value faster timing, more control over the process, more flexibility relative to a traditional IPO should consider the SPAC alternative. It helps if the company’s product is not too technologically complicated to be explained to non-expert public investors. It also helps if the SPAC has an experienced CFO and finance team that is ready for the rigors of being a publicly-traded company, including the need for detailed audited financials and timely financial reporting. I always say that a company needs to be ready not to go public but rather to be public.

There are risks. It is important to think through the alignment of incentives between SPAC sponsors and investors. SPAC sponsors require a meaningful fraction of equity. SPAC sponsors only get paid if the transaction happens so may be slow to walk away when conditions make it less favorable. Some SPAC sponsor team have not had expertise in the sector that the partner company operates in (that is, they may be financial teams without a sufficient grounding in technology or product). Some may not perform enough diligence.

Going public via SPAC won’t always be the optimal path for every company. But we think it is good for companies to have more than one path, and we think many will find the added control of the process can tilt the scales in favor of the SPAC approach.

14. The SPAC market has had ups and downs in recent years. How do you see its future evolving?

I think SPACs have historically been subject to swings; they have been in and out of fashion. In some periods, companies that went public via SPAC had stock that appreciated well for some period of time, even if the company wasn’t ready to be public. Then the image of SPACs went less positive, and we saw some quality companies going public and still experiencing a drop in share price. I think the market is leveling out, and investors are realizing that a SPAC business combination is not inherently good or bad but simply a tool. With good sponsors and management teams setting themselves up for good returns, and weaker sponsors and partners not seeing those returns.

Closing Thoughts

15. If you had to give one key piece of advice to investors looking to enter the venture space today, what would it be?

I would suggest that family offices thinking about becoming limited partner investors in venture funds:

• Determine the percentage of their wealth that can remain illiquid for a few years
• Diversify over three or more funds
• Use their venture fund investments as a way to gain expertise and familiarity with investing in startups, as a good intermediate step for investing directly in to startups themselves, and
• Most of all, see the current environment with lower valuations as an unusual opportunity

16. What’s next for you and Impact Venture Capital? Are there any exciting initiatives on the horizon?

We have launched our third seed fund and just made our first investment out of that fund. We have also launched a growth fund making later-stage investments in a subset of companies (from our Fund I portfolio) that have matured to the point where they don’t fit the seed-stage thesis that applied at the time of initial investment, and are getting closer to meaningful exits under shorter time horizons. And we are looking at partnering in a more meaningful way with other venture firms in addition to our corporate partners. We remain bullish on our thesis of investing in early-stage applied-AI startups in collaboration with tech corporates.

Some additional thoughts on the farther-future of AI

[From the shorter initial set of questions, question 6]

The Future of AI & Venture Capital – With AI advancing rapidly, where do you see the most promising opportunities for venture capital over the next five to ten years, and what advice would you give to AI startups looking to secure funding?

As mentioned earlier, AI already makes up over half of the business models for startups receiving venture funding; it is now the backbone of the majority of startups in the venture ecosystem. And now is breaking into subcategories.

For startups, I would advise them to focus not just on developing new AI tools but thinking about the business model the tool is being applied to. There are many startups succeeding by applying proven tools to new sectors or use cases. Also, try to avoid calling everything element of your business “AI”. Many software companies with new software are calling themselves “AI”, and it can become confusing. So be specific about what is new in your technology and how it is being applied.

Going further in the future, I think AI may not only represent the types of startups that will be funded, but can also be used by investors in ways that will change how investing is done across multiple asset classes. We are already at a point where information in publicly-traded securities is revealed and incorporated into prices very quickly, and I think it is much harder for a stock-picker to “add alpha” (that is, form a portfolio of similar risk that beats a stock market index) than it used to be. AI is helping public stock markets become more efficient. This will also now happen in other asset classes. It may become harder to ‘beat the market’ in trading bitcoin, or art, or cars as information is harder to keep private.

Even in markets that are based on developing and trading inside information, like private equity and venture capital, I anticipate these markets to become more efficient as well. There are some venture firms already using algorithms to drive their venture investment selection (such as Correlation Ventures in San Diego, whose algorithm makes decisions based on the track records of other investors in a startup). This will accelerate. I think in 50 years a new startup’s valuation may be unbiased (i.e. equally likely to turn out to be over or undervalued) and the market for trading in privately held stock may be as efficient as prices are today in publicly-traded stocks.

In past industrial revolutions, technology has displaced jobs, but this used to primarily blue collar jobs. In this AI revolution, many of the people displaced will be white collar educated professionals. These types of people may have wealth and influence and may not go quietly in to the unemployment line, so this could create social dynamics we have not seen before. And those displaced may include journalists and venture capitalists…

At Impact Venture Capital, we are excited about the potential for investing in seed-stage AI-oriented startups in collaboration with major tech corporates.