Flexport CEO Ryan Petersen on Logistics, AI and Fundraising

Ryan Petersen says fear of losing motivates him more than the thrill of winning, even though Flexport has grown far beyond his original expectations. He recalls an early financial model that projected $1 million in revenue and contrasts it with the much larger company he describes today.

Petersen offers a theory of venture-capital herding: investors may seek reassurance from peers because outcomes take years to measure and career incentives favor consensus. His comments are personal opinions about incentives and information sharing, not evidence of unlawful collusion by particular firms.

Stebbings and Petersen discuss whether financial security helps investors take unconventional risks. Petersen says living inexpensively in China in his twenties gave him room to experiment with businesses. The rent and living expenses he recalls describe his own circumstances at that time.

The two consider how money relates to ambition and self-worth. Petersen says he is motivated by doing large things, while Stebbings reflects on how closely investment work and financial outcomes can be connected. They present personal attitudes rather than a universal measure of success.

Petersen reports an approximately $450 million net-revenue run rate, compared with $350 million the prior year, and says he expects the business to approach breakeven. He also refers to more than $2 billion in revenue in another context. These are his company-reported figures and should not be treated as the same metric or independently audited results.

He offers two questions for evaluating rapid growth: is it driven by a temporary tactic that may stop working, and is the market large enough to sustain expansion? Petersen believes Flexport’s growth comes from ongoing sales work in a large market. His assessment and growth targets are not guarantees.

Petersen says he intends to take Flexport public after reaching stronger profitability, potentially in a few years. He does not know what valuation the market would assign. The timing and earnings milestones are aspirations, not a scheduled offering or financial forecast.

He says he prefers to focus on operating a profitable, growing business rather than trying to time an IPO window. The host asks about value concentrating in a few AI companies; Petersen says his main concern is retaining access to useful models.

Flexport handles different customer rules for shipment data, notifications and enterprise systems, Petersen explains. Traditional rules engines and automation can become difficult to maintain as variations accumulate. He believes agents may help handle more of those workflows, though the interview does not independently measure their accuracy.

He says the company uses models from multiple providers and may use open-weight models for appropriate tasks. A temporary account-limit message made him consider the risk of depending on an external provider, although he says the incident was an error rather than an intentional cutoff.

Petersen estimates Flexport’s model spending at about $5 million annually and says it has been growing. He expects a company to compare that cost with saved work or a better product. The amount is his estimate and the interview provides no independently verified return calculation.

He suggests routine automation might migrate to less expensive models while the most demanding development tasks continue to use frontier systems. His speculation that a provider might restrict customer access for other priorities is a hypothetical scenario, not evidence of an announced policy.

Stebbings questions whether AI productivity is overstated. Petersen says logistics contains many manual tasks, including passing documents and data between systems, which he believes offer opportunities for automation. He does not claim that every process is currently automated or that the gains apply equally across industries.

He says Flexport has nearly 2,000 employees and might remain around that size as it grows, with some work shifting from operations toward customer-facing roles. Headcount and job changes are his estimates and predictions, not a commitment or independently verified employment outcome.

The host suggests smaller numbers of highly capable salespeople could use AI to handle more work. Petersen agrees good tools can help, but argues global logistics still needs people to build and maintain relationships. They differ on the likely scale of that human effort.

Asked about forward-deployed teams, Petersen says enterprise sales can happen without them, but deeply embedded personnel can create strong customer relationships. He describes seeing logistics staff from another provider work within a large customer. The account is anecdotal and does not establish the customer’s decision-making process.

He recounts a separate customer win and offers his understanding of how a competitor’s staffing changes affected it. The individuals’ motives and the competitor’s actions are not independently established; the example illustrates Petersen’s view that relationships can matter.

Petersen says a largest customer spends around $150 million annually and that others spend tens of millions. He also gives broad estimates of logistics costs for physical-goods companies. These figures are company and personal estimates, not audited customer disclosures or universal benchmarks.

He says Flexport engineers can choose among coding tools and that the company also uses products integrated with its document environment. His account describes its own tool mix rather than a controlled comparison of providers.

In a hypothetical investment choice, Petersen favors Anthropic at equal valuations because he likes its enterprise business and team. He discloses that he has invested in Anthropic, recently rather than at its earliest stage. His preference is an interested personal opinion, not investment advice.

The speakers discuss whether open models may reduce demand for some proprietary models. Petersen says lower-cost intelligence could benefit users and that he does not know how the market will evolve. Their valuation and competitive implications are speculative.

Stebbings raises a concern about data exposure when using models associated with China. Petersen says he is less worried, but neither provides evidence that using a particular open-weight model sends data to a government. Actual data flows depend on deployment, provider and configuration.

Petersen reflects on having lived in China and believes economic interdependence matters in US–China relations. His views on geopolitical risks are personal judgments, not a prediction of policy or conflict. The panel’s brief comments do not establish the consequences of any particular deployment.

Turning to fundraising, Petersen recounts receiving an initial expression of interest on attractive terms, then approaching other investors in search of a better-known partner. He says the initial offer did not become a completed term sheet and the later process left him with different options. This is his account, not a finding about the other investor’s conduct.

He says he eventually told Founders Fund what had happened and discussed a lower valuation with different governance terms. According to Petersen, Peter Thiel offered a somewhat higher price than the one Petersen had mentioned. The interview does not provide transaction documents or the other parties’ accounts.

Petersen praises Thiel’s ability to think far ahead and recalls communicating quickly with him. Asked whether founders should always take the highest price, he says context matters and generic fundraising advice can be dangerous. The value of a partner, control terms and future needs can differ by company.

Stebbings argues that prominent investors can help with hiring and future financing. Petersen agrees that an investor’s reputation may provide a signal, while noting that employees often have to choose one employer at a time. This is their experience and opinion, not evidence of guaranteed fundraising success.

Petersen says Flexport shifted much of its workforce away from San Francisco because of costs, retention and the location of customers. He now thinks the leadership team may have become too dispersed and wants more colleagues in the same city. His headcount estimates are company-reported snapshots.

He strongly favors office work and describes difficulty concentrating at home with young children. His provocative characterization of remote work is his personal opinion; the interview provides no evidence that remote workers generally commit fraud or are uniformly less productive.

Petersen says Flexport’s baseline is in-person work and acknowledges that the company still relies on video across many offices. He sees recruiting from different geographies as another possible remote-work model. The effects of work arrangements vary by role, team and individual circumstances.

On hiring marketing leaders, Petersen argues that enterprise marketing requires originality and a willingness to experiment. He says employees may avoid unusual ideas if failure threatens their jobs, so founders need to create room for experimentation. His view is not a claim that conventional marketers cannot be effective.

He says customer events can work in business-to-business marketing and that not every event requires a senior executive. The examples reflect his experience; the interview does not provide a measured return for a specific campaign.

The host asks about human resources. Petersen says he values an HR leader who understands business goals and can maintain a two-way relationship with employees. His framing of the role is a management opinion, and employee interests and legal duties cannot be reduced to a single executive preference.

Petersen says Flexport has replaced some software but has not replaced the specific products Stebbings names. He expects the ability to build internal alternatives to affect vendor negotiations, while acknowledging that not every replacement would be worthwhile.

He describes using one internal replacement as a case study in procurement discussions and speculates about potential discounts. He admits some threats to replace products could be a bluff, because maintaining all those systems would divert engineers from the core business. His discount estimate is a prediction, not a negotiated result.

Petersen says security is a continuing concern in logistics and refers broadly to incidents in the sector. The interview does not independently verify claims about competitors or provide details sufficient to assess any named company’s security.

He describes making angel investments from an earlier profitable business and within his startup cohort. Petersen says he invested in some companies that succeeded and missed others, including Cruise. These are his recollections of investments and should not be taken as typical returns.

Petersen says he made money on early Bitcoin and invested early in Rippling. He and Stebbings discuss founder motivation, but their guesses about named individuals’ private feelings are not established facts. His personal investment results are not independently documented in the interview.

He says many investors passed on Flexport or stopped responding after meetings. Stebbings prefers giving direct feedback, while Petersen cites advice to distinguish an investor’s decision from the stated reason. Both are discussing personal fundraising experiences.

Petersen recalls leaving one meeting after repeated disagreement over the size of the logistics market. He says other investors were more interested in the founder and opportunity. His account of the meeting and market estimates are not independently verified.

He describes meeting Peter Thiel after an introduction and asking how a large logistics market fit a startup framework focused on smaller initial markets. Petersen says Thiel advised against applying the framework dogmatically and later invested. This is his account of their conversation.

From his angel portfolio, Petersen says a few unusually large outcomes can dominate many smaller returns. He marks speculative investments down mentally when he makes them. That is his approach to risk, not a recommendation or indication that such outcomes are typical.

The speakers discuss founder and investor incentives when a company might produce only a modest return. They agree that a founder’s personal mission and an investor’s portfolio calculations can differ. The appropriate decision depends on stakeholders, commitments and circumstances.

Petersen describes a fundraising rumor that reached other firms quickly. Stebbings says investors sometimes exchange information. Their anecdotes and speculation do not establish that all firms share confidential details or engage in unlawful coordination.

Petersen argues founders should be thoughtful about when and how they share operating metrics. His suggestion to emphasize favorable metrics is his opinion; it does not remove the need for accurate, non-misleading disclosures and compliance with applicable obligations.

He discusses the difficulty of fundraising while losing money and says narratives of setbacks and recovery can be compelling. For an enterprise company, he believes negative publicity can also be circulated by competitors. This is his view of business communication, not proof of any competitor’s conduct.

Petersen recounts a reported delivery problem and says the customer received a refund. He later acknowledges that more than one package may have been lost, so the interview does not establish a company-wide failure rate. His comments about reporting and competitors are his own account.

The conversation turns to marriage and family. Petersen shares a story about meeting his wife, a former journalist, and the professional boundaries they considered. Their experience is personal and does not establish a general rule for relationships or careers.

He recalls meeting SoftBank’s Masayoshi Son and says the firm led a $1 billion round. Petersen describes a suggestion to compete aggressively on price that he chose not to follow. The financing details and conversation are his account, not independently verified transaction records.

Petersen says Son’s network helped him seek feedback during the meeting. He describes public speaking and very short pitches as more nerve-racking than ordinary investor conversations. These stories illustrate his experience, not a prescribed way to raise capital.

He believes founder visibility and press appearances can help sales, although attributing a deal to one appearance is difficult. He says Flexport receives significant attention relative to its size; the interview does not independently measure its press share or ranking.

Asked about parenting, Petersen values a supportive partnership and describes the purpose he finds in family as well as work. His remarks are personal and do not prescribe when or whether anyone should have children.

In a quick-fire round, he says he has changed his mind about price competition. He now wants Flexport to pursue lower costs rather than rely solely on premium positioning. Whether it can become a low-cost leader is an objective, not an established result.

Petersen appreciates a supportive board that does not interfere excessively in operations. The host asks about sports sponsorship, and he mentions a team connected to a family association and a playful competitive idea. These are personal preferences, not a transaction announcement.

For a successful year, Petersen says he wants growth and a large majority of roughly 100 targeted AI workflows to deliver practical results. He says only a small subset was live at the time. The targets and expected savings are company goals, not verified future performance.