Coco CEO Daniel Singer on beating Starship in complex cities by partnering with Uber, DoorDash, and focusing ruthlessly on cost per mile

Summary

Coco CEO Daniel Singer makes a clear strategic argument: the delivery robot market is not one race with one winner. It splits by geography, and Coco is building for the harder half.

The Starship contrast

Starship has done well on college campuses and lower-complexity environments. Coco's bet is on dense urban markets — Los Angeles, Santa Monica, and similar cities — where sidewalk navigation is harder, regulation is messier, and order volume is highest. Singer argues that most delivery volume sits in big urban cities, which is exactly where competitors haven't fully cracked the problem.

The distribution model reinforces that focus. Coco is the only sidewalk robot company with active partnerships on both Uber and DoorDash, plus additional platform partners Singer says are coming soon. Not requiring customers to use a first-party app is the practical advantage: it gets volume without friction.

Cost structure as the core discipline

Singer frames Coco's philosophy around the Tesla-versus-Waymo divide in autonomy strategy. Waymo's sensor-heavy, high-capex approach makes it difficult to scale cheaply — Singer questions why, with Google's budget behind it, Waymo hasn't deployed tens of thousands of vehicles. Coco aligns closer to Tesla's camera-based approach: simpler hardware, lower per-unit cost, more units deployed across more locations.

The cost obsession extends beyond hardware. Singer says the team is focused on shaving cents per mile across every variable. But he adds a second dimension that often gets missed: reliability. A failed delivery — whether by human or robot — triggers refunds to the consumer and food-cost losses to the merchant. Getting to the 98th or 99th percentile on order accuracy is itself a cost lever, not just a service metric. Coco started with human-operated robots specifically to establish that reliability baseline before layering in autonomy gradually.

Timeline

Singer puts autonomous delivery becoming the default in major cities at one to two years, with the caveat that it's already the norm in some markets. In Santa Monica, he says, robot delivery is already what many residents expect. The rollout is market-by-market, not a single inflection point.

OpenAI partnership

Coco has announced a partnership with OpenAI. Singer declines to detail what they're building together but describes it as genuinely exciting. The implicit logic — raised in the conversation — is that OpenAI entering the physical delivery network business directly is a much larger strategic leap than it would be for a software competitor, which makes the partnership structurally less risky for Coco than comparable AI tie-ups might be for other sectors.

Tariffs and supply chain

Coco already has thousands of vehicles built and deployed in the US, which Singer says insulates the company somewhat from near-term tariff disruption. The broader robotics industry is more exposed, given that consumer electronics supply chains are heavily concentrated in Asia with few near-term alternatives. Singer floats capital gains relief on domestic manufacturing investment as a more targeted policy tool than a US sovereign wealth fund — a preference for incentives over government stock-picking.