Bastian Lehmann on Selling Postmates, Starting Again and the Future of AI

Harry Stebbings speaks with Bastian “Basti” Lehmann, co-founder and former CEO of Postmates and founder of TipTop. Lehmann recalls growing up in Germany, his curiosity about computers and a childhood shaped by his parents’ divorce. He credits his mother, who worked multiple jobs, with sustaining the family and inspiring his determination to support her.

Lehmann says he believes more in determination and hard work than luck, while acknowledging he has no definitive answer. As a teenager he wanted internet access but could not afford the dial-up charges. He recounts using calling-card credentials obtained by impersonating telephone-company representatives, describing it as the beginning of his computing career.

Lehmann describes Postmates as operating in wartime mode for much of its ten-year journey. He recalls a team committed to the company despite funding pressures, competitors and rumors. Rival fundraising was frightening even though Postmates itself raised roughly $900 million, according to his account.

He says the on-demand delivery market reached a point where capital for marketing and market share mattered greatly. Stebbings challenges the attractiveness of the model, citing low margins, density requirements and labor costs. Lehmann agrees delivery is operationally difficult but argues that the businesses had a path to profitability at sufficient scale; these are competing assessments, not an independent audit.

Founders Fund’s brand mattered when it led an early round, Lehmann says, because it provided a signal to others. But reputation and hype cannot rescue a company that fails to execute. In his recollection every fundraising round was difficult.

Uber CEO Dara Khosrowshahi called to discuss combining the companies. Lehmann recalls meeting him in a San Francisco park during the pandemic, after earlier discussions had not produced a deal. He believes Uber saw an opportunity to consolidate delivery, including regions where Postmates was strong while DoorDash led nationally.

Stebbings asks whether the sale was an emergency because Postmates was running out of cash. Lehmann rejects that characterization, saying the company had just under $100 million in cash and a negative gross margin in the single digits, and reached positive cash flow two or three quarters later. He says it was preparing to go public but decided a combination could be best in a market that needed consolidation; these are his account and interpretation.

Lehmann says he wanted to lead a public company and believed Postmates could continue independently, but set that ambition aside. The deal was announced at $2.65 billion; he says a fixed share exchange ratio without a collar meant the value rose to nearly $5 billion by closing as Uber’s share price increased. This refers to the transaction’s stock value at closing, not a guaranteed cash realization for each shareholder.

He recalls an initial offer around $2.2 billion and says he negotiated it upward. He prioritized deal certainty and the absence of a collar over pushing the headline price higher, and says he held almost all of his Uber shares at the time of the interview. His continuing stake is relevant to his favorable assessment of Uber.

Lehmann praises Khosrowshahi for giving trusted leaders autonomy, taking calculated risks and managing difficult company and regulatory issues. He says he initially mistook that style for insufficient involvement. Asked whether the Postmates acquisition had been written off, he says he saw no sign of the brand being abandoned and leaves any consolidation decision to Uber.

After the sale, Lehmann tried doing nothing and spent time with his family, but missed building products and working with engineers. Investing did not give him the same enjoyment. As a repeat founder who had returned capital, he found it easier to assemble a team, choose investors and navigate processes, though he says past success does not guarantee another.

He chose to work with Marc Andreessen, whose work he admired, despite having pitched Andreessen’s firm several times unsuccessfully with Postmates. Lehmann says Andreessen contacted him about his new project before the idea was fully formed. They discussed terms they considered reasonable rather than negotiating extensively; TipTop then had three co-founders and a problem area under exploration.

Lehmann admits too much capital can create complacency, while too little creates anxiety. TipTop raised enough, in his description, to test several hypotheses. He believes the product is promising but says the team is still testing market size and impact, and would be willing to move to another area if the evidence did not support continuing.

Financial security changes some practical decisions: Lehmann says he does not take a salary at the new company. But he believes ambition and the desire to prove oneself cannot simply be replaced by wealth. He says the team tries to remain humble, move quickly and assess problems without assuming prior success makes it better.

Stebbings asks how proceeds from a stock acquisition actually arrive. Lehmann describes a long process: signing, regulatory review, closing and transfer of shares through a custodian, followed by vesting of a small portion after a transition. He says there was no company-wide clawback to his recollection; his individual receipt and continued holding of stock are not a single cash payday.

During regulatory review, Postmates continued operating separately and working toward its targets because approval was not certain. The process felt anticlimactic, he says, and pandemic restrictions prevented a final celebration with the roughly 2,000 employees. He regrets that the team could not gather to mark the end of the journey together.

Lehmann is sharply critical of venture capital conformity, calling many investors followers of prevailing opinion. He argues that the best investors understand their influence on a company’s outcome is limited and avoid imposing themselves on founders. His view comes from his experience as a founder and investor; it is not a measured claim about all investors.

He recalls Founders Fund’s Brian Singerman telling him that an operator’s desire to change outcomes could make investing difficult. Lehmann interprets this as a philosophy of backing founders and staying out of the way. He says he applies a similar approach when he invests, while acknowledging that people can reasonably have different experiences with boards.

Lehmann advises founders to be wary when they must repeatedly educate an investor without building conviction. He worries that an investor may use the information to evaluate a competitor and that the relationship may become difficult when targets are missed. He acknowledges that his view is colored by his own experience and assumes capital can be found elsewhere.

He says he initially struggled to run board meetings because daily operations consumed his attention. Later he sent materials several days or a week in advance, held a dinner to discuss tensions and questions, and kept the formal meeting short and focused. The informal time also helped people reestablish rapport.

On AI, Lehmann argues that language-model computation was expensive at the time and imagines a home inference computer serving a personal model to other devices. He believes processing close to household data could support personalization and reduce cloud expense. This is a proposed architecture and cost hypothesis, not a demonstrated universal cost comparison.

He provocatively predicts that the phone has passed its peak importance, while acknowledging that phones will not disappear immediately and that a replacement may not yet exist. He is skeptical of the Vision Pro as then released, but imagines lightweight glasses with subtle information overlays. These are speculative product judgments made in 2024.

Lehmann expects incumbents to benefit from AI initially but predicts some startups will eventually become larger than today’s leading companies. Stebbings challenges his examples, arguing that current incumbents have unusually strong data, distribution and cash flow. Lehmann responds that established companies can weaken internally and that disruptive products may first look like toys; the disagreement remains unresolved.

Lehmann calls OpenAI a strong research company and expects it to be a major player, while saying success beyond ChatGPT remained to be seen. He is interested in specialized models and cites Figure’s robotics demonstrations as promising. These are his 2024 impressions and predictions, not independent evaluations of their products.

He argues that advancing AI and other technology is necessary to address challenges such as climate change and space exploration, and believes moving too slowly poses a greater danger than an “evil AI.” That is his personal risk judgment; the conversation does not establish the comparative safety of either path.

In the quick-fire discussion, Lehmann says mismatched expectations can create friction between investors and founders. He thanks Postmates co-founder Sean for years of support and says he would like more founder CEOs to stay in leadership. Looking toward 2034, he hopes to spend time at his children’s activities.