Detailed narrative
Call format — Q&A only, no prepared remarks
The call opened with IR (Weston Twigg) delivering only the safe-harbor and non-GAAP disclaimer before handing straight to Q&A; there were no scripted prepared remarks and no segment P&L walk. All financial color was elicited through analyst questions, so quantified disclosure was limited and largely qualitative, with management repeatedly pointing analysts to the shareholder letter and press release for figures.
Agentic AI and disintermediation defense
Tony Xu argued that top-of-funnel players (drawing analogies to Google Food Ordering, which ran ~8 years before shutdown, and Amazon's rise in product search) drive traffic but retain poorly, and that customers ultimately judge on the end-to-end experience — price, speed, accuracy, support.
He positioned DoorDash's proprietary, non-scrapable digital catalog of the physical world (billions of items, tens of millions per city) as the durable moat as agentic commerce grows, expecting AI platforms to remain partners driving a small, chosen share of traffic.
DoorDash is itself building agentic ordering experiences to lower discovery friction.
Autonomous delivery — DoorDash Dot
Ravi framed Dot as part of a broader autonomous delivery platform spanning land and air formats, launched in a couple of markets and scaling per the year's plan, targeting benefits in speed, quality and range of deliveries. Tony stressed the gap between shipping a demo vehicle and one that operates at scale under any condition — requiring hardening across autonomy, hardware, remote operations and city-level regulatory work — and said 2026 is about climbing that curve by doing the work in-house to be best-of-breed.
Global tech replatforming
DoorDash is investing several hundred million dollars into its platform, the largest component being a single global tech infrastructure stack replacing separate DoorDash/Wolt/Deliveroo stacks. Domain mapping/design is complete, production traffic is flowing and early benefits are appearing. Management framed dual payoffs: velocity (ship a feature once instead of three times) and quality (port best-of-breed logistics learnings between Europe's tighter, older cities and the US's larger geographies/catalogs).
All three legacy stacks run in parallel during transition, with redundant cost mostly through 2026.
Grocery and DashMart Fulfillment Services
Grocery, ~5 years in, is at record highs with DoorDash the volume share leader and capturing about 1 in 2 first-time grocery-delivery customers. Tony argued grocery should be as large as or larger than restaurants but the online experience isn't yet good enough versus shopping in-store; the fix is accuracy, affordability, easier basket building and ultimately solving fulfillment.
DashMart Fulfillment Services — an inventory-management/fulfillment setup live with a handful of grocery and retail partners — aims to deliver exact items with no substitutions/out-of-stocks, disruptive both to customers (positively) and to how retailers have long run supply chains, so it will be nailed before scaling.
Advertising
Tony said ads are at a record high and growing extremely fast, driven by cracking the code for SMBs, larger restaurant/retail advertisers and CPG advertisers, while balancing best-in-class return on ad spend against a deliberately lower ad load than other platforms.
Runway remains large including off-site/off-platform buying and in-store activity. Ravi framed incremental ad dollars like any unit-economics improvement — reinvested into the business to drive long-term free cash flow.
Europe competitive position
Against a premise that replatforming leaves DoorDash vulnerable to a more aggressive Uber in Northern Europe, Tony countered that the European position 'has never been stronger': Deliveroo is seeing its highest growth in four years and reaccelerating monthly, and Wolt is posting its highest share performance in each country it operates.
Ravi added that DoorDash is gaining share in the majority of markets and driving the majority of industry MAU growth.
AI-driven productivity and org structure
Roughly 2/3 of DoorDash's code is now AI-written, yielding faster feature/project/component delivery, but Tony noted the harder question is whether that translates to better customer outcomes and how team structures/workflows should change — a phase the whole industry is working through.
Priorities: get all teams onto a single tech stack, then make everyone (not just engineers) AI-capable. AI is also driving P&L benefits via agentic merchant onboarding, customer support and Dasher tools.