If you can't compete with AI, should you just deliver food? It's not that simple anymore.
```
As AI begins to replace white-collar jobs, many people joke, “Worst case, I’ll do food delivery”—but even this backup plan is quietly being blocked by robots.
Barclays released an in-depth research report on April 15, revealing that autonomous delivery robots and drones are reshaping the last mile of food delivery at an unstoppable pace. Today, the cost per order for human couriers is as high as $9 to $10, but this is facing direct competition from autonomous delivery technologies with a long-term goal of reducing cost to about $1 per order.
The research team visited Finland, Ireland, the US, Dubai and other places, witnessing Starship Technologies’ robots operating efficiently on the streets of Helsinki, and Meituan’s Keeta drones being deployed commercially in Dubai.
The Finnish market has become the clearest leading example—over 1,000 sidewalk delivery robots (SDRs) have achieved commercial-scale operations locally, and autonomous delivery accounts for nearly 10% of local GMV. Globally, private financing in the autonomous delivery sector will soar by more than 180% in 2025 vs. 2024; Zipline raised $800 million in 2026 (at a valuation of about $7.6 billion), and the industry’s strategic value is being systematically re-evaluated.
Currently, the penetration rate of autonomous delivery in global food delivery orders is still less than 1%, but Barclays predicts it will rise to about 2% by 2030 and may exceed 10% by 2035. Once penetration reaches 10% and each order saves an average of $4, autonomous delivery could unlock about $16 billion in annual profit space for global food delivery platforms.
In an optimistic scenario, as delivery costs fall and more price-sensitive consumers are attracted, the overall addressable market (TAM) for food delivery could expand by $40 to $80 billion in incremental GMV by 2035, raising the industry’s annual compound GMV growth rate from about 8% to roughly 9% between 2025 and 2035.
The Last Mile Cost Pain: Why Automation Is Inevitable
The food delivery industry operates on structurally thin margins, with delivery costs accounting for 40% to 45% of platform revenue per order, and labor accounting for about 80% of last-mile costs.
In mature markets like the US, the total cost per human courier order is about $9 to $10. With wage inflation, tight courier supply, and regulatory agencies in many jurisdictions pushing to reclassify gig workers as formal employees (such as the case of Glovo in Spain), this cost will only continue to rise.
Meanwhile, “tipping fatigue” is becoming more obvious in the US, further squeezing platform affordability.
Current tech-driven optimization methods—such as route planning and order bundling (with the best platforms achieving 30%-35% bundling rates)—are nearing their limits in low-density suburban markets, unable to fundamentally address labor cost issues.
The core logic of autonomous delivery is: shifting the cost structure from labor-dominated variable cost to hardware-depreciation-dominated fixed cost. As scale expands, marginal costs decrease, releasing operational leverage—a structural advantage the courier model can never achieve.
Two Technology Paths: Ground Robots vs Drones, Each with Their Arena
Based on fieldwork in Los Angeles, Helsinki, Dublin, and Dubai, Barclays deeply examined the two mainstream autonomous delivery technologies.
Sidewalk Delivery Robots (SDRs) are best suited for high-density urban areas, with a typical service range of 1-3 kilometers round trip, travel speed of 5-10 miles per hour, payload of 15-30 pounds, and 24/7 operation.
Representative companies include Starship Technologies (over 10 million deliveries, more than 300 deployment sites worldwide, a total travel distance over 22 million kilometers, and over 3,000 robots), and Serve Robotics (over 2,000 robots deployed in 20 US cities, with more than 1.8 million deliveries).
In leading markets like Finland, SDR delivery costs are $5-6 per order, $3-4 cheaper than human couriers. A single robot costs $9,000 to $18,000, with hardware depreciation as the largest cost factor (~55% of total SDR cost).
Drone delivery is more suitable for low-density suburban areas, with typical flights lasting 3-5 minutes, cruising at 65-80 meters altitude, payloads of 3-4 kilograms, delivering via winch to the ground (no landing required).
Representative companies include Manna Air Delivery (Ireland, completed over 60,000 deliveries in Dublin in 2025, aiming for ~2.5 million in 2026) and Meituan's Keeta Drone (operating about 65 routes in Shenzhen, Beijing, Shanghai, Hong Kong, and Dubai, with over 740,000 drone orders completed by the end of 2025, plus 5 million autonomous vehicle deliveries).
Current drone delivery costs are about $6-7 per order, and labor (loading/unloading and visual monitoring) still accounts for about 50% of total cost.
Finland currently provides the world's clearest commercial benchmark for autonomous delivery: about 10% of local GMV is fulfilled by over 1,000 SDRs—the highest known commercial proportion. In the long run, cost targets for SDR and drone delivery both converge to ~$1 per order, as publicly stated by Starship, Serve Robotics, and Manna Air Delivery.
How Much Can Be Saved Per Order
Barclays provides a single order cost comparison based on an AOV (average order value) of $30 in mature Western markets (benchmarking top-performing markets):
Traditional courier: total delivery cost ~$9.40/order, profit contribution ~$1.60/order (5% of AOV)
SDR (best market): total delivery cost ~$5.05/order, profit contribution ~$5.95/order (20% of AOV), saves $4.35/order vs courier
Drone (best market): total delivery cost ~$7.00/order, profit contribution ~$4.00/order (13% of AOV), saves $2.40/order vs courier
For SDR cost structure: hardware depreciation is about 10% (~$1.75/order), based on a robot costing $9,000, 4-year lifespan, 10-15% residual value, operating 6 days/week, 5 orders/day. Maintenance is about 3% (~$1/order), remote supervision labor ~3% (~$0.90/order).
DoorDash's model shows that for every 5 percentage point increase in autonomous delivery, FY28e EBITDA could see a mid- to high-single-digit percentage boost.
This leverage effect is because the marginal profit for autonomous delivery is higher—removing courier wages and tips lets platforms retain a larger share of each order’s income. The report also points out that future pricing model changes (from fixed fees to transaction-based pricing) will affect how much cost saving platforms can keep—a variable investors should track.
Potential to Unlock Hundreds of Billions Within a Decade
According to Barclays’ penetration forecast, the global value potential of autonomous delivery is estimated as follows:
Base scenario (10% penetration by 2035, ~$4 saved per order): annual global profit pool ~$16 billion. Sensitivity analysis: at 12% penetration, $6 saved per order, profit pool could reach ~$28 billion annually.
Upside scenario (TAM expansion): $40-80 billion incremental GMV by 2035, lifting industry GMV CAGR 2025-2035 from about 8% to around 9%.
The model assumes about 60% of automation savings are passed back to consumers via discounts and incentives, with a food delivery price elasticity coefficient (PED) of 1.5-2.0.
Scenario 1 (low AOV $15, high-frequency orders): ~$40 billion incremental GMV;
Scenario 2 (AOV $30 per order): ~$80 billion incremental GMV.
Five key drivers of TAM expansion:
Covers suburbs where courier economics don’t work (about 60% of Americans live in suburbs, but their share of orders is much lower, showing obvious gaps);
Lowers delivery fees to attract price-sensitive users;
24/7 delivery activates off-peak demand;
Expands to high-frequency categories adjacent to food (pharmacy, beauty, etc.);
Lower cost increases price elasticity for low-value categories (coffee, snacks).
By GMV, Barclays expects global autonomous delivery GMV to reach ~$160 billion by 2035—about $110 billion from SDR, and ~$35 billion from drones.
Seven Platform Automation Readiness Assessments: Who’s Out Front
Barclays built an “autonomous delivery readiness” framework using seven dimensions, concluding: DoorDash and Meituan are currently best-positioned to benefit in the near term, with Uber next; other platforms are mid- or long-term beneficiaries.
DoorDash is widely recognized as the leader.
The company has partnered with major autonomous delivery firms like Serve Robotics, Coco Robotics, and Wing, and launched its own robot “Dot” at its September 2025 R&D Day, building an “Autonomous Delivery Platform” (ADP)—an AI-dispatcher-like system that matches each delivery to the best method in real time based on speed, cost, and location. High labor cost pressure in the US, UK, and Nordics, a relatively friendly regulatory environment, and suitable urban layouts all drive DoorDash toward automation.
Uber uses an “asset-light collaboration” approach, with 1,000+ robots deployed in more than 10 cities and 7 partners. Its CFO has stated that current deployments are already “quite economical” and need no heavy upfront investment. Uber’s big challenge is maintaining a consistent business model and scaling pace across fragmented regulatory markets.
Meituan has a clear technical edge. Its Keeta drone has operated about 65 routes in Shenzhen, Beijing, Shanghai, Hong Kong, and Dubai, completing over 740,000 drone deliveries and 5 million autonomous vehicle deliveries as of end-2025.
However, this automation scale is still marginal compared to Meituan’s huge overall order volume, and its main challenge now is fierce competition in China’s food delivery market.
Delivery Hero, Prosus, Talabat, and Grab are classified as mid- to long-term beneficiaries, with automation still at the early pilot stage and not yet driving profits.
Seven Obstacles: Rising Penetration Won’t Be Smooth
Despite the rosy outlook, Barclays points out key structural barriers to scaling autonomous delivery remain.
Regulation is the biggest swing factor.
SDRs face highly fragmented local regulations in both the US and Europe (state-level in America, often down to municipal level in Europe). Drones face even stricter restrictions; BVLOS (Beyond Visual Line Of Sight) approval is limited to a few markets globally, but being first brings a strong edge.
The US FAA expects to finalize the “Part 108” regulatory framework by mid-2026, which would replace the current case-by-case waiver system with a national regime—a potential major catalyst for North America.
High capital requirements could slow the pace.
Most autonomous delivery companies are still private, unprofitable startups dependent on venture capital.
For example, Starship Technologies raised $50 million (Series C) in October 2025, reaching $280-325 million total; Manna Air Delivery raised $50 million (Series B) in March 2026, total over $110 million.
However, building the full stack—hardware manufacturing, maintenance, remote ops teams, and software—requires years of sustained capital, and current funding is not enough for massive global expansion.
Hybrid fleet management adds operational complexity.
For the foreseeable future, autonomous and human courier delivery will coexist. Platforms need multimodal dispatch: managing couriers, SDRs, and drones at once, routing orders across modes, handling bundling logic (robots don’t yet bundle), all without disrupting merchant workflow. Building these systems is a hidden barrier to scale.
Other headaches include: urban layouts (old European cities have narrow, complex streets bad for SDRs); uncertainty in automation pricing models (moving from fixed fees to variable pricing will affect profits); and cultivating public acceptance.
Cross-Sector Spillover Effects: Beverages, ESG, & Logistics
The impact of autonomous delivery goes far beyond food delivery.
Beverages: Barclays beverage team notes, greater convenience and extended delivery hours will systematically boost alcohol demand. DoorDash’s 2024 report shows alcohol delivery orders up 54% YoY, RTD cocktails up about 96%. The 24/7 reach of drones and SDRs covers peak demand at night and during sports events.
Sustainable investment: Autonomous delivery stands out in ESG.
Research shows delivery robots use over 97.5% less energy per kilometer than fuel vehicles, cutting carbon by 96%. Universal adoption could cut global carbon emissions by about 2.5% annually. Also, autonomy may reduce dependence on gig workers and labor compliance risk. For Meituan, for example, fully insuring couriers could add RMB 5.4 billion a year in cost (about 14% of 2026 EBITDA consensus estimates).
Logistics: Last-mile costs about 50% of e-commerce parcel logistics costs, so automation is also vital for traditional logistics companies—though longer distances and complexity mean scaling is harder than for food delivery.
~~~~~~~~~~~~~~~~~~~~~~~~
The above content is from Zhuifeng Trading Desk.
For a more detailed interpretation, including real-time research and field studies, please join [Zhuifeng Trading Desk ▪ Annual Membership]
Risk Warning & DisclaimerThe market is risky, investments should be cautious. This article does not constitute personalized investment advice and does not consider the individual investment goals, financial situation, or needs of any particular user. Users should consider whether any opinions, views, or conclusions in this article are suitable for their own circumstances. Investment is at your own risk. ```