Nothing summons an exec to the social media posting trenches quite like a venture capitalist drawing a misleading line chart about his company.
Last week, a16z declared that ridesharing is getting more expensive, pointing to new Gridwise Analytics data showing Uber fares rising roughly 20% since early 2024, while Lyft remained comparatively flat. Uber CFO Balaji Krishnamurthy barreled into the replies with the corporate equivalent of “well, actually”: aggregate prices, he argued, blur together the product and geography mix. Compare the same standard ride, at the same time and between the same places, and Uber tries to keep its price near Lyft’s. The apparent premium comes from customers buying more premium products and Uber expanding into suburban markets where trips run longer. CEO Dara Khosrowshahi then entered the chat to announce that “The Professor is back in the house!”
As fun as it is to watch these titans of industry catfight, Uber’s CFO makes a great point. Gridwise’s Rideshare Trend Report is measuring the rides users actually took, not a controlled basket of identical routes and quotes. Uber pricing is indeed on the rise, but that’s because riders are generally choosing “better” and longer mobility options.
Getting a Lyft upmarket
From Q2 ‘25 to Q2 ‘26, the average observed Uber fare rose 8.8%, from $22.63 to $24.62, while Lyft’s rose 6.2%: from $18.67 to $19.83. Uber’s average ride ended the quarter about 24% more expensive than one on Lyft. Its median fare also rose faster, up 10.3% versus Lyft’s 7.1%.
But those averages hide quite the array of products, from a shared ride around the corner to a private Escalade towards a distant airport. Premium products — including Black, Comfort, Priority and XL — represented 20.1% of Uber’s Q2 rides, more than twice that of Lyft’s. Uber’s premium share grew 15.5% year over year and 43.5% over two years. Lyft is moving upmarket too, but from a much lower base.
That is the report’s juiciest useful takeaway: this isn’t just a story about the same ol’ ride getting pricier; this is about the basket of rides changing. Uber has built, and persuaded more customers to choose, a larger menu of roomier and faster and fancier trips. Balaji himself says that Uber’s mix has shifted toward journeys longer than 15 km as it grows its footprints in suburban and “sparser” markets.
Mix doesn’t explain everything. Platform fees rose on both apps: Lyft’s average fee jumped 26.4% YoY to $3.11, while Uber’s rose 14% to $5.30. This makes “premiumization” look like a convenient arrangement for everyone: riders choose a nicer product, drivers receive more per trip and the platform has a larger fare from which to collect its take.
Drivers get a raise, probably…
Those higher fares aren’t just vanishing into the app store’s maw. Average driver gross pay per completed trip rose 7.6% YoY to $15.78, while base pay rose 6.9% to $12.22. Gross pay per working hour increased 7.3% to $24.41, and average quarterly gross pay reached a record $4,119, despite drivers working 2.1% fewer hours.
Those are meaningful gains, but they of course come with a few big caveats. Gridwise’s figures come before fuel, maintenance, depreciation, insurance, taxes, vehicle payments, etc. Insurance is also becoming less of a headwind. Gridwise’s estimated per-trip insurance figure fell 5.1% at Uber and 5.5% at Lyft from Q1 25-26.
More money, more waiting
The pay story comes with one little bump in the road. Trips completed per working hour slipped 0.6% to 1.68, while utilization — the share of observed working time spent on passenger trips, measured from request to drop-off — fell from 62.99% to 62.40%, the lowest level since 2021. Drivers earned more per trip and per hour, but spent a slightly larger share of their time deadheading. There could be a number of factors at play here: increasing driver supply, spatial mismatches, the aforementioned change in product mix or even the rise of robotics.
Speaking of robots
The question on everyone’s mind is, what sort of impacts are autonomous vehicles having on human ridehail drivers? Gridwise compared human-driver outcomes between Q4 2024 and Q4 2025 in five cities where robotaxis are active. The results are noisy enough to frustrate both the boosters and the doomers.
LA posted the toughest numbers for flesh-bound drivers: quarterly gross pay fell 20%, trips per hour dropped 9.7% and utilization swooned by 7.8%. San Francisco’s drivers completed 4% fewer trips per hour, yet quarterly gross pay managed to rise 5.7%. Atlanta and Austin recorded higher gross pay per trip and per hour despite fewer trips per hour, while Phoenix saw insignificant changes.
As they say, correlation is not causation. Southern California has over a hundred thousand ridehail drivers, while fewer than 1,000 robotaxis ply its roads, at least for the time being. It does not seem the specter of mass job losses at the hands of our autonomous overlords is yet upon us…
In fact, all things considered, the report paints a picture of a surprisingly healthy, but perhaps increasingly segmented, TNC market. Customers are paying more, drivers are grossing more and platforms are collecting higher fees. And Uber’s larger premium mix helps explain why its average fare sits so far above Lyft’s.
In a way, a16z and Balaji can both claim their preferred chart is “correct.” Rides are getting more expensive. But that’s because they’re getting “better.” The looming question is whether customers will keep shelling out for more and more upgrades, and what happens to drivers once even more riders get a taste of that truly upmarket robotaxi experience.
HOT INDUSTRY NEWS & GOSSIP
Micro-cars, macro storage: As more and more clever startups unveil sub-car scale vehicles, one familiar question has been “but where do I put my stuff?” Check out this inventive new design from Belgium-based Any, with 120 liters of internal storage, about 1/3 what you’d find in a compact car. Put another way: that’s over four cubic feet, good for about three standard carry-on rolling suitcases. Meanwhile, Liux just unveiled a new upmarket microcar.
Delivery bots get schooled: After Starship’s abrupt exit from the U.S. college market, it was open question as to which other delivery bot companies would pick up those contracts. As predicted, Avride and Robot.com are the biggest gainers, with the former adding 21 campuses and the latter signing a seven-year contract with food service co. Sodexo.
Your self-driving hackney, guv’nah! Uber and Wayve have officially gone live in the United Kingdom, taking 15 Ford Mustang Mach-Es live in London. The cars still have safety drivers on board; they may be there a while, given that TfL is behind schedule on finalizing the rules for AV deployments.
Next stop on the NATO express: Latvia has finished up major upgrades to Riga Central Stations, a key component of the transnational Rail Baltica program. That mega-project is linking Latvia, Lithuania and Estonia to Poland (and thus to the rest of the E.U.) via higher-speed standard gauge rail, augmenting those countries’ older Russian-gauge networks.
Who watches the watchmen? The anti-Flock saga continues, with Midwestern flair. Grand Rapids, Michigan has ordered the city’s ALPR systems go through the Public Safety Committee review and City Commission approval process, in a move meant to improve transparency and trust.
There will be no flying cars… Surprise, surprise, Pivotal — the flying car company backed by Larry Page — just lost its CEO. As much as everyone that loved the Jetson’s wants to envision a world where we all whisk around in our own li’l flying pod, people can already barely drive when sticking to just two-dimensions. Add in fuel, noise, insurance, PUDO concerns, etc — and you can see why this is apt to remain a pipe dream.
State capacity gets the good: Paris just okayed an extension of the three year old T10 tram line. This project is more metro than streetcar, with the 3.5km long line running in a tunnel, and set to cost €730M. That’s about $390M per mile: expensive for France, but dirt cheap for America.
Modern Delivery, summer break quick hits: Instacart & Chain’s Cart Classics social media slam dunk, an interview with IC’s Director and Head of Instacart Ads Studio. DoorDash earns Alliance for Audited Media incrementality certification, after Circana partnership. World Market launches on Instacart. More feces in the food supply! California passes bill mandating price transparency on cargo ZEVs. Temu invests in local fulfillment. Chipotle heads to Asia. Popwheels completes 500,000th battery swap. Walmart furthers its restaurant delivery ambitions, as it adds Dunkin to its in-app options.
A few good links: Tesla FSD seems to really love train tracks. California Gig Workers Union expects certification. NJ looks to up wages for microtransit drivers. Minneapolis and Seattle seek curb management technology vendor. More e-bikes head to Berlin. Things are looking up at Greyhound. Thessaloniki, Greece opens metro extension. IL inaugurates Northern Illinois Transit Authority to better integrate regional transport. Waymo goes live in Denver, Tampa and SD. Trump’s aluminum tariffs to hamper domestic auto industry. Uber abruptly pulls out of Nigeria, leaving some riders and passengers in the lurch mid-ride. An eager The Curbivore reader shares their take comparing Seoul and L.A. Results from AIWaysion’s 15-month NYC curb activity pilot. Waymo’s full fleet size revealed.
See you in Austin!
- Jonah Bliss & The Curbivore Crew
Premiumization has definitely played a role in Uber's ability to increase average rideshare prices and maintain a sizable price premium over Lyft. But these gains reflect Uber's superior algorithmic price discrimination capabilities -- i.e., the ability to target the right price to the right customer at the right time. After all, the same vehicle can be deployed on any given trip as an UberX or Comfort Ride, so the trick is figuring out which customers are more likely to opt for the promise of a better ride. The same is true for Priority Rides, which may or may not guarantee faster pickup times (you'll never know!), so the trick again is Uber's superior ability to target such rides to customers most likely to opt in.
Uber would have us believe that while their superior price discrimination algorithms on both sides of its marketplace convey an unbeatable asymmetric information advantage over riders, drivers, and competitors, THEY DON'T EXPLOIT THAT ADVANTAGE. Is their denial of using personal information in setting prices and pay credible?