Confessions of a Delivery App Coder: Why Your Delivery App Wants Its Drivers Poor
An anonymous user on reddit posted insider details into how delivery apps harm their drivers and lower their pay
It’s been an open secret in the world of food delivery that the “system” is rigged against driver pay. We pay for priority, yet our food arrives cold. We tip generously, yet drivers still seem to struggle.
Early on January 2nd, an anonymous Reddit poster claiming to be a backend engineer for a major delivery platform spilled the beans on a myriad of predatory algorithms that aim to keep driver pay as low as possible, while maximizing profit for food delivery companies.
You guys always suspect the algorithms are rigged against you, but the reality is actually so much more depressing than the conspiracy theories…They talk about these people like they are resource nodes in a video game, not fathers and mothers trying to pay rent.
Most of us have clicked that $2.99 “Priority” button to get our food delivered a little quicker. According to the whistleblower, that button does absolutely nothing to change the logistics of your order. In the backend code, it simply flips a “boolean flag” (a true/false marker) that the dispatching logic ignores.
Even more cynical was a reported A/B test conducted by the company. Instead of speeding up priority orders, management allegedly purposely delayed standard orders by 10 minutes. The goal? To make the “Priority” service feel faster by comparison. It wasn’t about efficiency; it was about manufactured frustration.
The poster claims it “generated millions in pure profit just by making the standard service worse, not by making the premium service better.”
Arguably the most heartbreaking part of the leak involves how the algorithm treats its most dedicated workers. Drivers aren’t seen as people; in the database schemas, they are literally labeled as “human assets” or “resource nodes.”
The system reportedly tracks a “Desperation Score” based on a driver’s acceptance behavior.
The Trap: If a driver is desperate enough to accept every low-paying $3 order, the algorithm tags them as “High Desperation.”
The Consequence: Once tagged, the system stops showing them high-paying orders. The logic is chilling: why pay a desperate person $15 for a run when you know they’ll do it for $6? High-paying orders are instead saved for “casual” drivers to “hook” them into the app, while full-time workers are squeezed for every penny.
In response to new labor laws, many apps have added a “Driver Benefit Fee” or “Regulatory Response Fee.” Most customers assume this money helps provide healthcare or insurance for the person at their door.
The whistleblower claims this is a total fabrication. Instead, that money reportedly flows into an internal cost center for “Policy Defense.” Essentially, customers are unknowingly funding the high-powered lawyers and lobbyists who work to prevent drivers from unionizing or gaining employee rights.
These companies pay nothing while their customer base directly funds lobbying efforts to keep drivers disenfranchised.
After previous lawsuits, delivery apps stopped “stealing” tips directly. However, the leak describes a more sophisticated version: Predictive Modeling.
The algorithm predicts how much a customer is likely to tip based on their history.
If you are a high tipper, the app lowers the “base pay” it offers the driver (e.g., $2).
If you are a non-tipper, the app raises the base pay (e.g., $8) just to ensure the food gets moved.
The result? Your $10 tip isn’t “extra” money for the driver—it’s a subsidy for the corporation. You are paying the driver’s wages so the company doesn’t have to.
The whistleblower ended their post with a sobering sentiment: “I can’t sleep at night knowing I helped build this machine.”
As consumers, we often focus on the convenience of the interface. But behind the sleek UI is a backend built on “Desperation Scores” and psychological manipulation.
The next time you order, remember: the person delivering your food isn’t just fighting traffic; they’re fighting an algorithm designed to keep them down.
I drove for Doordash for most of 2025, until I was kicked out due to a suspended license. The alleged whistleblower's observations line up with my long-held suspicion that wage theft has been coded into the system. I imagine they use deliberate obscurities in their code to escape detection. It's a long standing "worst practice" well known to coders of shady e-commerce and porn sites.
Remember when the gig economy was going to make everyone rich?