Corporate Venturing Insider #100 with Jeff Bussgang of HBS & Flybridge Capital final video

Background & Journey

Host: What was your journey into corporate venturing, entrepreneurship, and venture capital?

Jeff: I started as a computer nerd in the 70s and 80s, teaching myself programming on early personal computers. After studying computer science at Harvard, I worked at BCG for a couple of years to learn business fundamentals, then returned to Harvard for my MBA. In 1995, I joined the internet software company Open Market when it was small. Within a year, we went public, reaching a couple of billion in market cap. Over five years, I served in various executive roles, including VP of product, marketing, professional services, and business development.

After leaving, I co-founded a company called Upromise, and in 2002, launched Flybridge with friends who spun out of venture firms like Greylock. We saw an opportunity to create a venture capital firm focused on early-stage investing in the Northeast.

Consulting Background

Host: How did your consulting experience at BCG help you as an investor?

Jeff: The glib response I give my HBS students is that "I learned a little bit of strategy and a lot of PowerPoint." But the real answer is that consulting helped me think about value creation at the core. Having both technology enthusiasm and business fundamentals has been valuable throughout my entrepreneurial and venture capital career.

Host: I saw a video where you had provocative thoughts about choosing entrepreneurship versus consulting.

Jeff: In today's environment, I think people would be crazy to go back into consulting after experiencing entrepreneurship or an MBA program. In consulting, you don't make consequential decisions - you just advise and generate pretty reports. As a builder, either as a joiner or founder, you get to make real decisions. By becoming an entrepreneur, you learn much faster and more deeply than as a consultant on the sidelines.

Entrepreneurship Reality

Host: You make entrepreneurship sound easy, but most people say it's very hard.

Jeff: I've had a charmed journey with two successful startups, but it's like a duck that looks calm on the water while paddling furiously underneath. We had many ups and downs - nearly running out of money, product failures, having to reboot our codebase a month before IPO, missing quarters, and dealing with failed acquisitions.

In my venture career spanning 22 years with hundreds of investments, I fail more often than I succeed. Entrepreneurship is incredibly hard. It's like playing at a poker table where you're not just playing one hand but multiple hands over a 20-30-year career, hoping a few play out well.

As an investor, I'm always respectful that entrepreneurs don't have that portfolio dynamic - they're doing one company at a time, with everything in one basket. My job is to help them nurture and protect that basket.

The Experimentation Machine Book

Host: Tell us about your book, which Eric Ries called "the perfect guide for AI era founders."

Jeff: The book is called "The Experimentation Machine: Finding Product Market Fit in the Age of AI." It builds on the mindset Eric Ries and Steve Blank started with lean startup methodology, now flourishing with modern AI tools that allow entrepreneurs to experiment more effectively and efficiently.

It's based on my work with portfolio companies at Flybridge (which is now exclusively AI-focused), as well as my research at Harvard Business School in collaboration with OpenAI, Salesforce AI, Microsoft, and Google. The book covers how founders can leverage these capabilities to find product-market fit more rapidly. It's available at jeffbussgang.com or on Amazon.

Investment Thesis Development

Host: How do you define and build an investment thesis?

Jeff: I teach this at Harvard Business School in my Venture Capital Journey class. An investment thesis begins with a hypothesis about the future - ideally, a non-obvious one. In 1994-95, believing the internet would become a place for secure enterprise business was controversial, though it seems obvious now.

In 2025, the job of a venture capitalist is to develop non-obvious hypotheses about the future. As Howard Marks describes in his 2x2 matrix, on one dimension you're either right or wrong, and on the other you're either consensus or non-consensus. Being right and consensus don’t create alpha because it's already priced in. You need to find non-obvious opportunities.

After developing your hypothesis, you go deep and narrow on a manifestation that you believe will yield tremendous value. This requires strategic thinking about value creation, competitive moats, and pockets of margin. Then you develop criteria against which you evaluate startups, looking for ones that could seize that opportunity consistent with your hypothesis.

Publishing Investment Theses

Host: If your investment thesis gives you an advantage by identifying something non-obvious, why share it publicly?

Jeff: When I first started blogging as a VC 20 years ago, it was itself non-obvious that VCs should be transparent. As a former entrepreneur, I cared about educating other entrepreneurs and sharing what I was learning from behind the curtain.

We publish our theses to share ideas with entrepreneurs and attract those who align with our thinking. If an entrepreneur reads our thesis and says, "That's me! I need to talk to this person," that's successful outreach. It establishes credibility and helps us connect with founders who share similar beliefs about the future.

Host: It seems like it also helps filter for entrepreneurs who resonate with your thesis.

Jeff: Exactly. And it forces us to be crisp about our thinking. It's one thing to casually mention interest in agentic AI in an interview, but another to write four pages informed by primary research with buyers, entrepreneurs, PhDs, and academics. It's a forcing function for us to be more precise about our thesis.

Flybridge's Current Investment Thesis

Host: Could you share your investment thesis for our audience?

Jeff: At a high level, we believe every company will be an AI-forward company. As I say in my book, AI won't replace founders anytime soon, but founders who use AI will replace those who don't. We want to invest in "AI-forward founders" - not just companies building AI products, but founders operating like "10x founders" using AI tools.

Second, we believe the infrastructure layer is being rapidly commoditized with deepseek, open-source models, and declining AI compute prices. This follows the pattern of previous tech cycles where infrastructure companies dominate early (like Cisco in internet days or ARM in mobile), before application companies create massive value.

We're focused on:

• AI-forward vertical SaaS

• AI-forward tooling and development environments

• AI-forward horizontal applications replacing functions like HR and customer service

• Applications targeting $200,000/year jobs like legal, accounting, and consulting

• Infrastructure for managing thousands of AI agents in the workforce

We believe every company will hire more AI agents than humans, so we're investing in management tools, authentication, security, orchestration, feedback, and quality control for large AI agent workforces.

Future of Corporate Reporting

Host: Can you imagine companies reporting how many AI agents they have in 10 years, similar to how they report employee headcount today?

Jeff: That's provocative. In my book, I wonder if we've reached peak employment in the Magnificent 7 tech companies, as they've grown revenue 20-40% while keeping headcount flat. Companies might report revenue per employee as an increasingly important metric - can you build a billion-dollar company with 100 employees? 50? 10? As Sam Altman said, maybe even one employee?

Companies might need to publicly disclose their AI tech stack because it's becoming a fundamental part of their competitive advantage - how many agents they're using, their investment in AI, and what capabilities they're deploying in production.

Host: Cisco used to measure revenue per employee. While we may have reached peak employment, we certainly haven't reached peak revenue per employee.

Jeff: Totally agree. The efficiency ratios in our portfolio are extraordinary. We have one company that reached $40 million in revenue in two years with just 20 employees. Another service that thousands of customers use with thousands of AI agents and only 20 humans. And we're just at the beginning - only two years since the launch of ChatGPT. We're very early in this development curve.

Advice for Corporate VCs

Host: Having worked with CVCs, what's your advice on value creation for corporate venture capitalists?

Jeff: There's a distinct segmentation among CVCs. The real test is whether you can be more than just the wire that sends money. Can you deliver true market insights, expertise in a particular segment, and most importantly, business development opportunities and revenue contracts? If you can provide these elements of value-add, you're a top-quartile CVC in my book.

CVC Challenges and Pitfalls

Host: Let's discuss some challenges with CVCs. Without naming names, what bad practices have you observed that CVCs should avoid?

Jeff: There's a fundamental difference in approach. Institutional investors like us have a single focus: "Can we make money on this investment?" Every decision revolves around whether it creates equity value.

Less effective CVCs sometimes allow corporate interests to influence their decision-making. They might block financing, slow down important decisions, or hold companies hostage by requesting special deals or renegotiations. For startups, you can't afford delays in critical decisions.

Host: Could you elaborate on specific examples of problematic behaviors?

Jeff: As investors, we recognize that not all startups succeed. If a company isn't performing well, we believe in "exiting with honor." For example, if market conditions suggest selling the company to get our money back or even taking a small loss is the best path, we'll recommend that to founders rather than prolonging the pain.

Some CVCs, however, might push to continue for another year or two. Perhaps the individual is moving to a new role and doesn't want the loss on their watch, or they're avoiding delivering bad news to superiors. This leads to behaviors like blocking reasonable transactions.

I've seen cases where a company was doing well enough to sell for $150-200 million - not a rocket ship, but a good outcome where founders and early investors would make money. The CVC, which came in at a high price, might only get its money back, so it threatened to block the sale unless given a 1.25x return. This is incredibly shortsighted. In one case, they eventually settled for 1.1x, but nobody on that cap table will ever work with them again. When we see them talking to founders, we share this story and advise against working with them.

This is a reputation business with multiple rounds at the table. Short-sighted behavior leads to long-term poor ramifications.

Stability and Reputation

Host: VCs and CVCs may look similar, but are quite different. Flybridge has stability with team members staying for many years, whereas CVCs often have more turnover. How does this affect reputation, which takes "10 years to build and 10 minutes to lose"?

Jeff: There is churn in venture capital, too, especially as firms grow larger. However, I've been at Flybridge for 22 years, my co-founder Chip Hazard has been here 22 years, and my partner Jesse Middleton, who leads our New York office, has been with us for nine years.

It's frustrating for founders when they have to re-educate a new board representative every few years because the previous champion has been promoted or moved aside. This makes us cautious about bringing CVCs into the cap table and boardroom - we want stability and institutional memory.

Board Seats vs. Observer Status

Host: Let's discuss fiduciary duty for board members versus observers. Could you explain why this matters and how entrepreneurs and CVCs should approach board observer seats?

Jeff: As board members, our fiduciary duty is to act in the best interests of all shareholders, not just one class or our own interests. Sometimes we may take actions against our employer's interests, but for the benefit of all shareholders.

As venture capitalists, we have dual fiduciary roles - to our LPs and to our portfolio companies. Rarely do they conflict, but sometimes we might pass on a follow-on investment because we think it's throwing good money after bad.

Board observers have no fiduciary duty and no vote. They attend meetings and access information, but don't participate in voting. Voting becomes critical in split decisions regarding M&A transactions, financings, or CEO changes.

I often advise entrepreneurs taking CVC capital to offer observer seats rather than board seats. There are times when board observers should be excused for "board only" sessions, particularly when discussing competitors to the CVC, potential partnerships, or M&A transactions. You need these firewalls between the CVC (who may act in their employer's interests) and the board (who must act in all shareholders' interests).

Host: What would you recommend when a CVC board member needs to discuss engaging with a competitor?

Jeff: The CVC should proactively recuse itself, and the entrepreneur should proactively assert that there should be a recusal. This is beyond best practice - it's a must-do. While CVCs rarely object to recusal requests, they don't always proactively recognize when they should step aside.

When a CVC might be an acquisition candidate, we want them outside the room during acquisition discussions so we can develop strategy and negotiation tactics without conflicts.

CVC Best Practices

Host: After discussing pitfalls, could you share examples of CVCs doing things right?

Jeff: The best CVCs act like team members - almost like a part-time VP of Business Development. They actively ask how they can help open doors, make introductions, and champion the startup within their ecosystem.

I had one CVC investor who would go unannounced to their CEO's office and say, "You've got to meet with Jeff. What he's working on is important." He was shameless but charmingly aggressive, which is exactly what you want. He got me in the room, and we ended up signing a company-making deal.

It's too easy to send an email, get rejected, and give up. A good VP of Business Development or sales rep with a quota wouldn't take no for an answer - they'd find ways around obstacles to help their startup succeed. That's what we want from CVCs in our cap table and boardroom.

Host: Cash from CVCs is good, but cash from customers, thanks to CVC introductions, is 10x better.

Decision-Making Best Practices

Host: What decision-making practices from VCs could benefit CVCs?

Jeff: At Flybridge, our decision-making style is "high conviction and non-consensus." You need thoroughly considered investments, but unanimity isn't required. We can have dissent in the partnership - someone might question the founder, product, market, or valuation.

As long as the champion thoroughly considers feedback, gets more data, and iterates, we can make non-consensus decisions with high conviction. The non-consensus decisions tend to be the best ones.

Second, nothing kills deals like time. The most competitive deals don't hang around long. VC firms are incredibly agile - we can meet a founder on Friday, deliver a term sheet over the weekend, and close before Monday morning. I rarely see CVCs execute with that speed, but it's necessary.

This requires having a prepared mind, investment thesis, and focused due diligence on the 2-3 key issues. We make decisions with imperfect information but with conviction. Many CVCs have bureaucracy, processes, or extensive box-checking that would prevent deals from happening in institutional VC.

Host: The "high conviction, non-consensus" approach is challenging for CVCs since corporations typically favor consensus. Could you elaborate on the "prepared mind" concept that enables fast decision-making?

Jeff: The term comes from Charlie Munger of Berkshire Hathaway. The idea is that if you do your homework on a sector and develop an investment thesis with clear criteria, you're ready when opportunities arise.

When you meet an entrepreneur who checks all your boxes in a 45-minute meeting, you can immediately recognize the fit. This also allows for efficient due diligence - rather than "boiling the ocean," you can focus on the 1-3 salient issues that separate success from failure.

Host: Your investment thesis ends with criteria that help you quickly identify good fits. But what about startups outside your thesis that are compelling? How do you balance your prepared mind discipline against FOMO?

Jeff: My partner has a saying: "We don't need to invest in every great deal. We just need the deals we invest in to be great."

That said, we remain open to outlier ideas. Sometimes we'll take time to understand a new industry or opportunity. For example, with Blitz Engineering (which you and I invested in together), the company applies to the auto and battery industries. We had a prepared mind for cloud-based machine learning and testing, but we weren't experts in batteries and EV equipment.

It took extra time to validate the founder's hypothesis about the future, the market size, and whether automakers would outsource to a cloud-based service provider. We needed multiple meetings to understand the industry.

Host: With Blitz, they perfectly met our criteria because we have expertise in batteries and electronic components from our mothership. That's our superpower.

Jeff: What makes us deviate from our criteria is often an exceptional team. At the seed stage, the team has an enormous impact on outcomes. If we have high conviction that the team members are outliers, we might be more accommodating on other dimensions.

We also believe amazing markets with strong tailwinds can deliver great outcomes even with mediocre execution. If we have high conviction in the market, we might compromise on price or traction, though rarely on team.

For example, we recently invested in an AI application company where we paid a higher price than we might normally because the team was outstanding and the market had exceptional tailwinds. History shows this approach can work well - our early investments in MongoDB and FalconX faced similar pricing questions, but both became multi-billion-dollar companies.

Host: So if you're going to invest in what becomes a fund returner (25-50x), paying up is reasonable because it's still an outlier.

Jeff: Exactly. Fund returners have become bigger over time. When we first invested in MongoDB, we thought the market might be $8 billion, and we could build a company worth $1-2 billion. Today, MongoDB has over $2 billion in revenue and a $20 billion valuation. Outcomes are larger than ever before, with many companies worth tens or hundreds of billions. This creates opportunities for both entrepreneurs and investors to take more risks with massive market tailwinds.

Creating a Culture of Constructive Dissent

Host: How do you create a culture where dissent is welcome, especially for junior team members? Most corporations favor consensus.

Jeff: My Harvard Business School colleague Amy Edmondson uses the concept of "psychological safety" - an environment where people feel safe to take risks. Google has implemented this carefully in its practices.

Creating psychological safety means attacking ideas but not people, encouraging well-considered risks even when they don't succeed, having constructive disagreement, avoiding defensiveness, and making the best possible decisions together.

Advice for New CVCs

Host: What recommendations would you give to CVCs who are just getting started?

Jeff: When I transitioned from entrepreneur to investor, a mentor observed that sometimes your strongest relationships develop across firms rather than within your own, particularly with co-investors on boards.

I'd encourage new CVCs to seek out mentorship from respected investors in their portfolio companies. This is a craft that takes decades to learn, with learning cycles of 8-10 years. Finding seasoned investors on your cap tables or in boardrooms and seeking their guidance is invaluable.

Host: I love that insight. This is essentially an apprenticeship that takes time to master, and learning from others' mistakes and failures is incredibly helpful.

Jeff: I agree.

Host: Thank you for your time and these golden nuggets. For the audience, remember Jeff's book will be valuable for understanding AI investments and selecting the