Uber One Doesn't Reward Riders. It Quietly Subsidizes Drivers.

Based on the research ofZhang, Cavusoglu and Raghunathan, "Membership Plans for Sharing Economy Platforms," Information Systems Research, 2026

Every major sharing-economy platform has converged on the same subscription design: the rider or shopper pays a monthly fee, the driver or gig worker pays nothing. Uber One, DoorDash's DashPass, Instacart+, and Grubhub+ all charge the consumer and hand the provider nothing but, implicitly, a more committed set of customers. This convergence looks like an obvious best practice. A working paper forthcoming in Information Systems Research by Haozhao Zhang, Huseyin Cavusoglu, and Srinivasan Raghunathan says it is closer to an accident of habit, and that whether membership belongs on the consumer side, the provider side, both, or neither depends on a variable most platforms never name out loud: which side of the marketplace is currently scarce.

The Model: Membership as a Scarcity Instrument, Not a Perk

Sharing-economy platforms do not look like traditional retailers on either side of their ledger. A retailer's supply is captive and its marginal cost is close to fixed; a ride-hailing or delivery platform's supply is a rotating pool of independent workers who decide, hour by hour, whether to log on, and its marginal cost and revenue swing with real-time surge pricing on both the consumer and the provider side. Zhang, Cavusoglu, and Raghunathan build a model of exactly that environment and ask where a membership plan should sit once you take self-scheduling and variable marginal costs seriously.

Their answer breaks the symmetry that "just subscribe your best customers" assumes. A consumer-side-only plan is the platform's best move when expected supply potential is high relative to demand; a provider-side-only plan is best when expected demand potential is high relative to supply. The two plans also make money in different ways: a consumer-side plan earns its keep mostly through the membership fee itself, while a provider-side plan earns its keep through the operational profit that a steadier, more committed pool of providers makes possible. Offering both sides a plan at once, the paper finds, only beats offering no plan at all when neither side's potential is particularly large; when the platform can lean hard on scarcity in either direction, a one-sided plan already captures more value than trying to lock in everyone at once.

The sharpest result, though, is distributional rather than financial. Introducing a membership plan on only one side systematically harms the side that bought it and benefits the side that did not. A consumer subscription is not a discount the platform extends to riders out of generosity; per the model, it is a mechanism whose gains flow disproportionately to providers, funded in part by the very subscribers who signed up expecting to come out ahead. Flip the plan to the provider side and the direction of harm flips with it. There is no version of a one-sided plan in this model that is a pure win for the side holding the membership card.

What Uber One and DashPass Actually Are, Read Against the Model

Line up the paper's prediction against the market and the pattern is almost too clean. Uber One, launched in November 2021, bundles 5 percent off eligible rides, $0 delivery fees on qualifying Uber Eats and grocery orders, and, notably, "priority service with top-rated drivers" for $9.99 a month or $99.99 a year. DashPass offers unlimited $0-delivery-fee orders over a threshold and reduced service fees for a comparable monthly price. Instacart+ does the same for grocery delivery, adding a family-sharing feature so one subscription can cover a household. Every one of these is a consumer-side-only plan, in a set of markets where, after years of aggressive driver and shopper recruitment in most metro areas, supply potential is plausibly the more abundant side of the equation the paper describes.

That detail in the Uber One benefit list is worth sitting with: members get routed toward "top-rated drivers." That is not a footnote. It is the mechanism chain from the paper made visible in a product spec. The platform collects a fixed fee from consumers, and part of what that fee buys is preferential matching that concentrates demand on a favored slice of the driver pool, exactly the kind of provider-side benefit the model says should show up once a consumer-side plan is in place. DoorDash's own program makes the redistribution even more explicit: active Dashers get DashPass itself, the consumer product, free or discounted, provided they complete at least one delivery a month. That is a platform visibly trying to return some of the value a consumer-side plan pulls toward providers, by literally handing providers the consumer membership as a perk.

The Missing Half of the Market: Why Provider-Side Fees Barely Exist

What is striking once you go looking is how lopsided the real market already is. Consumer-side membership fees are everywhere in the sharing economy: Uber One, DashPass, Instacart+, Grubhub+, all charging riders or shoppers a recurring fee. A provider paying a platform a subscription fee for better treatment is, by contrast, essentially absent from the market. What exists instead on the provider side is free, performance-gated loyalty. Lyft Rewards sorts drivers into Silver, Gold, Platinum, and Elite tiers based on monthly ride earnings, unlocking gas cashback, roadside assistance, and scheduling perks, at no cost to the driver. Amazon Flex runs a comparable points system, where drivers accumulate points per delivery and per scheduled block, unlocking a debit card with cashback on fuel and Whole Foods purchases and the ability to reserve preferred shifts, again free to join.

A one-sided membership plan does not add value to a marketplace so much as move it, and right now almost every platform has moved it toward the side that pays.

This asymmetry is itself a finding, even though no one designed an experiment to produce it. If the paper's threshold logic is right, a genuine provider-side membership, one where drivers pay for better matching, higher effective wages, or priority scheduling, should be the platform's best move whenever demand potential runs high relative to supply: a hot market, a surge period, a city where riders consistently outnumber available drivers. Those conditions exist regularly, seasonally and geographically, across ride-hailing and delivery. Yet no major platform charges providers a subscription fee for it. Either every platform's supply is genuinely abundant everywhere all the time, which is not consistent with how often riders see surge pricing and long wait times, or platforms have simply defaulted to the consumer-fee template because it is the one every subscription business outside the gig economy already uses, and nobody has run the scarcity check the model says the decision actually requires.

The Decision Rule Hiding Inside a Subscription Button

The practical error the paper exposes is not that Uber, DoorDash, and Instacart chose consumer-side plans. In markets where supply genuinely is the abundant side, that is plausibly the value-maximizing choice, exactly as the model predicts. The error is treating the choice of side as a solved problem rather than a live one that should move with local market conditions. A platform expanding into a new city with a thin driver base, or entering a seasonal window where demand reliably outstrips supply, is looking at the mirror-image case the paper describes, one where a provider-side plan, not another consumer perk, is the profit-maximizing and possibly the fairer design. Defaulting to "add another Uber One tier" in that setting is optimizing for the wrong side of the ledger.

There is also a welfare dimension operators should not wave away as someone else's problem. If a one-sided plan harms the side that buys it, a platform that keeps stacking consumer-side benefits onto a market where supply is actually the tighter constraint is not just leaving money unclaimed; it is asking its subscribers to fund gains for the side that is already, structurally, in the stronger bargaining position. Riders paying $9.99 a month for "priority service with top-rated drivers" are, on this account, not purchasing a service upgrade so much as financing better matching for a driver pool the platform needed to court all along. That may still be a reasonable business decision. It stops being a defensible one the moment a platform tells itself the subscription is simply a gift to loyal customers rather than a redistribution mechanism with a direction, a magnitude, and a losing side.