Martin Casado on Building Markets and Rethinking Venture Capital

Martin Casado traces his route from computational physics and national-lab research to a systems PhD at Stanford. He founded Nicira from that research in 2007; during the 2008 downturn, he says, Mark Andreessen and Ben Horowitz helped save it. VMware later bought the company, and Casado ran its business before joining Andreessen Horowitz to focus on infrastructure.

His operating experience shapes his view of boards. He describes many investor directors as well intentioned but too distant from product-market fit to contribute much, and others as frustrated operators who try to drive a company that is not theirs. He names Horowitz as an unusually effective board member.

Casado is unsure whether operating experience helps an investor pick or win deals and thinks it may hurt. It can, however, build empathy: he recalls personally paying payroll, nearly shutting down, repeatedly seeking capital and nearly losing his board's confidence. The key is to help the founder without trying to run the business.

Asked how he handles numerous boards, he points to his work habits and team. He says that, while building an organization of about 4,000 people and hundreds of millions in revenue run-rate, he often worked 100-hour weeks; he views board work against that operating experience.

He credits junior colleagues, including Jennifer Li on data and Matt Bornstein on AI, as well as specialist recruiting, sales and marketing teams. In his account, a16z's support staff allow partners to bring expertise that an individual GP cannot match, and he believes he sometimes receives credit that belongs to the team.

Casado says his interest in infrastructure continues outside board meetings: he joined a Discord community with Tim Chen to discuss the field. He would rather investigate technology, category creation and market positioning than spend all his time chasing the next deal.

He argues that the private technology capital market has expanded enormously from two decades earlier, while much of venture's generalist model comes from that smaller era. His estimate that the earlier market was one-fiftieth its current size is his own characterization, not a sourced measurement in the interview.

His proposed response is specialization, larger platforms and funds spanning stages and asset types. He predicts a barbell: focused boutiques can prosper in a niche, while scaled firms can serve companies from seed through public markets; firms caught between may face pressure from both sides.

Casado contrasts venture's founder-oriented, innovation-focused identity with what he sees as the priorities of public markets, lenders and private equity. That is his broad critique, not a demonstration that all other investors disregard innovation. As an a16z general partner, he has a direct interest in the model he promotes.

He wants technology investors to support founders throughout a company's life, not only until a later financier takes over. He recalls a private-equity investor questioning why founder leadership mattered; Casado believes founders can retain credibility and help companies navigate transformations.

Harry Stebbings challenges the multistage model: if a board member's firm may lead the next round, a founder might withhold bad news in order to keep selling. Casado says that would indicate a broken relationship and notes that at a16z different fund teams decide at different stages.

Casado says he mainly invests at seed, Series A and Series B, and has pitched the firm's growth team to back a company. He considers independent decision-making between funds important. Stebbings also worries that enormous funds could reduce venture returns toward private-equity levels.

Casado answers that investors can offer different funds for different risk and return profiles: debt, early-stage equity and late-stage equity need not promise the same results. He says his own focus is less on extracting every possible gain from each dollar than on gaining exposure to leading innovators.

He believes technological progress contributes to human welfare and expects substantial value to accrue to innovators, though the interview supplies no controlled evidence for those broad claims. He describes a portfolio of leading software companies as an index-like way to participate in that opportunity, not a guarantee of individual investment returns.

Stebbings raises a criticism that Casado is insufficiently price-sensitive. Casado disputes it, saying he generally invests at the low end of competitive prices and has not done the billion-dollar Series B rounds implied by the criticism. He asks for a specific deal to test the claim.

Asked about dbt's entry price, Casado says the figure is not public but characterizes it as low. He distinguishes the later high prices achieved by portfolio companies from the prices he himself paid. The interview does not disclose the deal terms, so neither account can be independently assessed from it.

His investing method has changed. Casado says he once believed he could determine whether a single early-stage company was a good investment, but now considers the problem underdetermined, especially for first-time founders creating a category. He sees comparison among a group of companies in the same space as more tractable, while still uncertain.

He studies an area deeply and compares its participants rather than claiming to predict a company in isolation. He credits Chris Dixon with another lesson: look for fields that attract exceptional founders, whose collective choices he considers a useful signal of emerging opportunity.

That research includes the market's direction, the players and the path from technology to adoption. Casado says conversations with Databricks CEO Ali Ghodsi often turn to category creation and go-to-market strategy rather than the technology alone.

He calls the sustained work of creating a market “market annealing.” Drawing on his experience with software-defined networking and network virtualization, he says a startup may have to educate and reshape a market for years; selling an improvement into an existing category presents a different challenge.

On timing, he compares market decisions with Go: the number of possible states is too large for neat rules. He offers a tentative heuristic rather than a formula: building against a shrinking macro trend makes an already hard task harder.

For a new category, the necessary work depends on market readiness and a startup's staying power. Casado says some companies create a market through sheer persistence over a decade. Better technology in an established category may instead bring questions of churn and pricing.

Storytelling matters, he argues, both for educating a market and for leading a team. He points to HashiCorp's Dave McJannet and Aviatrix's Steve Mullaney as leaders who can articulate why a company matters and what future it is trying to create.

His suggested exercise is to write a clear, deeply held vision and use it consistently in recruiting, culture, early sales, marketing and conversations with analysts. The vision need not specify every step, but it should explain the destination and why the team is pursuing it.

Stebbings asks whether spending cuts in a tougher funding environment undermine category creation. Casado rejects the trade-off as inevitable: a company must size itself for the market and available funding while continuing to do the work needed to create its category.

He warns that a blanket hiring freeze can prevent managers from moving people, addressing performance or allocating resources to changing needs. Instead, management should revisit the operating plan and decide where to invest or cut; reductions in spending do not automatically require layoffs.

In the 2022 downturn, Casado recommends preparing bear, middle and bull cases, even if none is immediately enacted. His advice is shaped by 2008, when he says Nicira could not raise money, he paid salaries from his bank account and a team of ten stayed roughly the same size for three years.

A bear case might model a substantial revenue decline and a long period without new financing; a middle case can use public-company or comparable data; a bull case considers stable or accelerating demand. The purpose is to have a response ready if a quarter falls far short, not to pretend any scenario is certain.

Asked how much to cut and whether to make layoffs once, he says boards should first agree on the level of risk and an operating plan designed to keep the company viable. Some boards may choose to accelerate hiring in a downturn. The appropriate response, in his view, follows from the plan rather than a reflexive percentage target.

Casado admits he made mistakes as a young board member by overfitting his own experience. He gave advice as though he were running another founder's company and treated his own successful path as a general template, even though he had disliked the same behavior when he was a founder.

He advises junior investors to contribute information that founders have less time to gather: map competitors, meet companies and synthesize market research. Unless they truly know an area, he says, they should avoid prescribing company strategy and leave the founder responsible for operating decisions.

He argues that the industry undervalues junior partners' market knowledge, even when criticism of inexperienced operating advice is fair. That distinction also influences his own team: investment debate is shared, and junior colleagues can have significant influence on decisions.

For difficult decisions with many unknowns, Casado cites Andy Grove's argument for vigorous debate over days, weeks or months. He says his team spends much of its time discussing infrastructure and mapping markets so that it can act when a transaction moves quickly.

Stebbings asks how a fast deal allows time for deliberation. Casado says the work must largely happen before the deal arrives; if the team has not studied a space, that may be the reason it passes. He believes research and direct conversations cannot be replaced by manufactured deal momentum.

The host asks how junior colleagues can safely disagree with a seasoned operator. Casado acknowledges being wrong often and says longer tenure has improved shared language and trust. He also admits that his own strong opinions make it important to keep watching whether others can challenge him.

Engineering taught him to rely on techniques that have worked and avoid repeatedly failed ideas. Investing requires a different mindset, he says, because successful companies can emerge after several earlier iterations of an idea; he cites search and social networking as examples.

He credits Marc Andreessen with showing him a more optimistic view of technology's future. Stebbings worries about “tourist” VCs who prioritize visibility and social media over innovation; Casado disagrees that the profession was better in the past.

Casado recalls earlier investors who, in his experience as a founder, were distracted and generalist. He sees more disciplined specialists now, while acknowledging that a decade-long boom had also attracted opportunists. His expectation in 2022 was that the downturn would drive some of the latter away.

Discussing a16z's growth, he says separate investing arms, including bio, crypto and venture, gained autonomy and could build platforms for their own markets. He also says scale creates practical difficulties: distributed teams must clarify processes, budgets and where profit and loss responsibility sits.

Asked where he could improve as a leader, Casado returns to credit. He says he is often named for deals and research in which others did more of the work. Giving the infrastructure team fair recognition matters for individual careers and for understanding where the value actually comes from.

For young investors facing a run of disappointing investments, he recommends identifying what they can control and articulating an approach they believe in. He identifies two recurring hazards: emotions that distort follow-on decisions and principal-agent incentives that can lead team members to favor their own careers.

His own approach is to understand spaces where strong founders are active and choose among promising companies, then judge his decisions against that process. Stebbings asks whether such an approach risks confirmation bias: an investor might find evidence that supports a thesis that is itself wrong.

Casado distinguishes his approach from a grand theory of how a market must develop. He says founders help identify the space, while investors may bring a broader view from meeting many companies and talking with customers. He fears a sweeping narrative can too easily be retrofitted to the evidence.

Looking ten years ahead from the 2022 interview, Casado hopes more capital will support innovation throughout company life, including after an IPO. He believes technology investment can produce predictable returns at scale; this is his forecast and investment philosophy, not an established result.

He does not predict that traditional finance will disappear. He expects venture-style capital to take some territory as markets grow, while acknowledging that predictable returns also serve purposes such as retirement savings.

For a favorite book, he chooses The WEIRDest People in the World, describing its thesis about Western institutions, psychology and cooperation among strangers. His account is a discussion of the book's argument, not independent proof of its historical or physiological claims.

Asked what he learned from Andreessen, Casado returns to a belief that technology can improve the future. He also says he expects the cloud to disaggregate, and wishes he had started his board career with less hubris and more realistic expectations of venture investors' abilities.

He calls the crossover investment model potentially sound but argues that financial investors may struggle with early-stage markets they do not understand. Among his misses, he says his team had not invested in the cloud security posture management category, citing companies such as Lacework, Orca and Wiz.

Casado's broader conclusion is that the industry can miss opportunities when it acts as though capital deployment must remain constrained by an earlier, smaller market. That claim reflects his outlook at the time, including his belief that the opportunity was large even amid the downturn.

The best investment advice he names is to trust founders and follow the founder network. He singles out angel investor Ram Shriram for the work and value he says Shriram brings to boards, despite his wealth and prominence.

Unable to identify his most recent publicly announced investment during the quick-fire round, Casado instead describes where he hopes to be in ten years: continuing to promote product- and founder-focused investing as a larger part of finance. Stebbings closes by thanking him for the conversation.