Only 4 Uber Rides Were Booked by AI Agents in Its Busiest Week
In short
- Uber says only four rides were booked through AI agents in its busiest October week, out of about 320 million trips and orders.
- Uber Eats got around 500 agent orders in the same week. DoorDash reports a few hundred in total.
- The money angle: do not pay for an agent-checkout project yet. Fix your own site speed and checkout first.
Everyone says AI agents will do our shopping. Uber just published the real number. It is tiny.
What happened
On its third-quarter earnings call on November 4, Uber said that during its busiest week of October, the week of Halloween, only four rides were booked through an AI agent. Uber Eats saw about 500 orders from agents. Total volume that week was roughly 320 million trips and orders across both apps.
DoorDash told a similar story: a few hundred agent orders so far, and most customers still use the app. Uber CEO Dara Khosrowshahi called it very early with very limited traction. He also said every AI company is talking to Uber, and he expects more than 10 major assistants over time.
There is a fun detail. Uber says booking a ride in its app takes about 50 seconds, and reordering a favorite meal takes about 90. So an agent does not even save much time yet.
Do the math: 500 agent orders out of 320 million is about 0.00016%. If your shop gets 2,000 orders a month and agents behaved like Uber’s, you would get roughly 0.003 agent orders a month. That is zero. Now say a developer wants €1,500 (my assumption) to build an agent-ready checkout. You would pay €1,500 to serve a channel that sends you nothing today.
How to use it
Do not panic and do not build. Spend the effort where the orders really come from.
First, make your normal checkout faster. Uber’s own app needs 50 seconds, so beat that and you beat the agent. Second, keep product data clean: clear titles, prices, stock and delivery times. Agents read structured data, and so do Google and your customers. This work pays off either way. Third, set a reminder to check your analytics every quarter for visits from AI agents. When they pass 1% of orders, then it is time to invest.
What it means
The agent shopping story is early, not dead. Big platforms want to be ready because one big assistant could change habits fast. But for a small seller, the order is clear: customers first, agents later. Marketing decks promise a wave. The data shows a few drops.
Who loses? Anyone selling agent-readiness packages today. Who wins? Sellers who kept their money for ads and stock.
Ivan’s Take
Not worth building for right now. I run GiantParts and I would rather put €1,500 into better product photos and faster pages than into a channel with four rides a week at Uber’s scale. I will watch the numbers each quarter and move when agents reach a real share.
Source: TechCrunch
If it doesn’t save you time or money, you won’t see it here
- Facts, not hypeEvery price and number is checked against the official source
- Always the real priceYou see what it costs before you click
- I say when it’s weakEven when it’s OpenAI or Google
