A sweeping new investigation has revealed that grocery-delivery platform Instacart has reportedly been using AI-powered pricing tools to charge different customers wildly different prices for the same items — even when shopping from the same store. The findings raise fresh concerns about fairness, transparency, and the unchecked use of algorithmic pricing in everyday consumer services.
What the Investigation Found
- Independent researchers working with Groundwork Collaborative, Consumer Reports and More Perfect Union tracked 437 shoppers across four U.S. cities who simultaneously added identical grocery items to their Instacart carts from the same stores.
- The experiment found that about 74% of items tested displayed varying prices depending on the shopper — not the store. In some cases, the same product appeared at up to five different price points at once.
- On average, the price difference between lowest and highest prices for an item was 13%; for some items, the gap was as large as 23%.
- In concrete terms: one test basket of groceries from a store in Seattle cost some shoppers about US $114, while others were charged as much as US $124 — a nearly 9% difference for the same shopping list.
- If such pricing discrepancies are widespread rather than occasional experiments, some households could be paying hundreds — or over a thousand — extra per year compared with others buying identical items under identical conditions.
How Instacart and Its Technology Fit In
At the heart of these pricing variations is an AI-driven revenue optimization tool supplied to retailers — a tool originally developed by Eversight, a company acquired by Instacart in 2022. According to the investigation, that system allows retailers to run real-time “price experiments,” shaping what individual users see based on factors such as supply, demand, competitive pressures, or retailer-defined strategy.
In response to the revelations, Instacart acknowledged that price experiments occurred but insisted they were “limited, short-term and randomized,” and that only a small number of retail partners participated. The company maintained that most customers see “standard prices” and that any variability was not based on user-specific personal or demographic data.
Some major retailers reportedly involved in the tests, such as Target and Costco, responded by distancing themselves from the experiments — stating publicly that they do not control which prices appear on Instacart and do not share individualized price data with the platform.
Why This Matters — and What’s At Stake
- Consumer fairness and trust: Many users assume that if two people buy the same item at the same store, the price should be the same — just as in a physical grocery checkout line. These findings may shatter that expectation.
- Budget unpredictability: Price swings up to 23% make it difficult for households, especially low- and fixed-income ones, to plan grocery budgets reliably.
- Opaque algorithmic practices: Because pricing experiments can take place behind the scenes and vary per user, it becomes nearly impossible for consumers to detect or contest unfair pricing.
- Regulatory and ethical scrutiny: As algorithmic pricing becomes more common in retail and grocery delivery, regulators may view such practices as a form of price discrimination — especially if personal data or opaque targeting is involved. The issue dovetails with broader debates around fairness in algorithmic commerce.
Responses and What Might Happen Next
In light of the public outcry and media attention, several actions seem likely over the coming months:
- More retailers may exit or halt participation in dynamic-pricing experiments on platforms like Instacart, especially if public pressure mounts. Indeed, some have already reportedly pulled back.
- Consumer-protection groups and possibly regulators may call for greater transparency about when algorithmic pricing is used, which users are affected, and what data drives price variation. Some states have already begun introducing laws aimed at “algorithmic pricing disclosure.” Public awareness around “surge-style pricing” in everyday retail may increase, prompting users to price-compare across platforms or favor stores committed to stable, transparent pricing.
ADMT Convenience, but at What Cost?
Online grocery platforms promised convenience, speed, and easy access. But behind the slick apps and one-click ordering lies a new kind of uncertainty — not just over availability or delivery times, but over what you’ll actually pay for the same items someone else buys just moments before or after you.
The findings from Groundwork Collaborative, Consumer Reports, and More Perfect Union suggest that algorithmic “experiments” on pricing may be quietly re-shaping the meaning of fairness in retail. Whether the public — or regulators — will accept this quietly dynamic pricing model remains to be seen.
