Overview
Group E Current Event - Instacart and Dynamic PricingBackground and Event Summary
In 2022, Instacart acquired a tech company called Eversight and began using its AI to run real-time pricing experiments on its grocery delivery app, testing how shoppers responded to different price points. For years, most customers had no idea this was happening. That changed in December 2025, when Consumer Reports and the Groundwork Collaborative published an investigation where 437 volunteer shoppers in four cities bought identical grocery baskets from the same stores at the same time. Every volunteer ended up inside a pricing experiment. About 74% of items were priced differently depending on the shopper, with an average gap of 13% between the lowest and highest price and some items varying by as much as 23%.
One of the more striking findings involved how discounts were displayed. Shoppers saw different "original" prices for the same item, even when everyone paid the same sale price, making a deal appear bigger or smaller based on which group a shopper had been randomly placed in. Consumer Reports estimated that families consistently sorted into higher-priced groups could pay up to $1,200 more per year. The FTC opened an investigation shortly after, and on December 22, 2025, Instacart announced it was ending retailer-led AI pricing tests immediately. The company rejected the "dynamic pricing" label, framing the tests as short-term randomized experiments rather than personalized pricing. Either way, the tactics connect directly to several of the decision-making biases we've covered in class.
Anchoring Bias
A key concept from Daniel Kahneman’s Thinking, Fast and Slow is the anchoring effect, which explains how people rely heavily on an initial reference point when making decisions. This bias is especially relevant in today’s digital marketplace, including Instacart’s use of dynamic pricing AI. Instacart adjusts prices in real time based on factors like demand, location, and shopping behavior. A common feature is the display of a comparison price, such as an “in-store price,” next to the Instacart price. This comparison acts as an anchor that shapes how consumers perceive value. An example could be an in-store priced item at $4.99, but the Instacart price is $5.79. Even if the shopper does not typically track prices, the $4.99 becomes the mental benchmark. Kahneman explains that System 1 thinking quickly uses this anchor to judge the higher price. Instead of evaluating whether the markup is reasonable, the shopper perceives the increase as minor “only $0.80 more.” As a result, System 2 thinking is less likely to engage.
This bias has important implications. First, it influences perceived fairness as long as the price remains close to the anchor, consumers are more likely to accept it. Second, it reduces price sensitivity, discouraging comparison with other retailers. Third, it promotes quicker purchasing decisions, driven by instinct rather than careful analysis. AI-driven pricing raises concerns about transparency and consumer awareness. While dynamic pricing increases efficiency, it also allows companies to subtly influence decisions through cognitive biases. Instacart’s pricing model demonstrates that consumer choices are not purely rational but shaped by how information is presented. Overall, anchoring bias shows how digital platforms can guide behavior, reinforcing Kahneman’s insight that human judgment is often influenced by quick, intuitive thinking.
Framing Effect
Instacart used the framing effect to deflect the pricing structure criticism targeted at the AI pricing model. Historically, retailers will often fluctuate the prices of certain items in grocery stores to keep products moving off the shelves. However, these prices are consistent for all consumers and only vary based on time the product spends at the store, making the comparison that Intacart was implying inaccurate.
Additionally, Instacart showcased their AI dynamic tool as a way to save money and that on average most customers end up saving. Unfortunately, according to a Consumer Reports study in 2025, customers averaged paying 7% more across the twenty most popular food items. Another false claim was found after the results of this study were released. Previously Instacart has used standard market conditions, and retail partners are responsible for any perceived price hike shown in specific products. Conversely, Target, a major retailer associated with Instacart, denied having any formal business relationship with the company. Soon after this statement was made, Instacart began pulling back on their pricing experiment.
Loss Aversion
In Thinking, Fast and Slow by Daniel Kahneman’s, he describes the bias of loss aversion, which refers to how psychologically the feeling of pain and loss is experienced twice as intensely as the joy of pleasure or a triumph, which results in a fear or aversion to risk taking decisions. This manifests in a variety of areas (ie. social status, opportunities and even sentimental objects, etc.), though most often in decisions relating to money. As a result, this cognitive bias heavily informs and influences the way consumers view price differentiations and fluctuations, particularly when they think they might miss out on a deal or pay more than others.
A recently conducted investigation by the nonprofit organization, Consumer Reports and Groundwork Collaborative, found that “identical grocery items on Instacart could differ in price by as much as 23% from one customer to the next.” (CBS News, 7) In some cases, it has been estimated that a family of four’s estimated grocery costs have increased by $1,200 a year, as a result. Instacart, like many companies, weaponized and manipulated the cognitive bias of loss aversion, intending to give the illusion of larger savings to customers. Similar to the case that took Amazon to court in October 2025 over “fictious pricing” used during the summer Prime Day Sale, Instacart was found, per the report, to be showing customers differing “original” prices, giving some sales the appearance of larger savings. Ultimately, Instacart leveraged and manipulated this bias in tandem with unethical business practices to influence consumer choices under the guise of providing fictious sales designed to provoke System 1 thinking, giving the platform an increase in profit at the cost of the trust and relationship of their customers.
Conclusion
Instacart's use of AI-driven dynamic pricing ultimately shows how impactful behavioral economics can be in the outcome of consumer decisions. Concepts such as the anchoring effect, framing effect, and loss of aversion help us see how customers perceive value. Evidence throughout the paper shows that pricing can vary as much as 23% on identical items between guests. This data point shows the extent of perceived discounts as well as customer manipulation. While the company may come ahead in profits, these tactics raise huge concerns about customer fairness. This leads us to the ultimate question: Is personalized pricing in this way concerned with smart marketing or overall unethical? It is more important now than ever that consumers closely pay attention to their surroundings, so they are not lost amongst technologies' newest innovations. Companies and consumers should hold each other accountable for a fair and reasonable shopping experience.
Please answer the following:
1. Is the experience of shopping online more prone to cognitive bias than the in-store experience?
2. In practice, can a rational shopper protect themselves from deceit when a store displays an "original price" and then a "sales price"? How much comparison price shopping is realistic given the relative price differences (between the real "original price" and the displayed "original price")?
References
[1]
Consumer Reports. (2025, December). Instacart's AI-enabled pricing experiments may be inflating your grocery bill. https://www.consumerreports.org/money/questionable...
[2]
Groundwork Collaborative. (2025, December). Same cart, different price: Instacart's price experiments cost families at checkout. https://groundworkcollaborative.org/work/instacart...
[3]
Instacart. (n.d.). Eversight: Optimize your pricing with AI. https://www.instacart.com/company/retailer-platfor...
[4]
Palmer, A. (2025, December 22). Instacart ends AI-driven pricing tests that pushed up costs for some shoppers. CNBC. https://www.cnbc.com/2025/12/22/instacart-ai-pricing-tests-increased-costs.htmlLinks to an external site.
[5]
Natureindex. 2026. Psychological Pricing Strategies in Consumer Behavior. https://www.nature.com/nature-index/topics/l4/psychological-pricing-strategies-in-consumer-behaviorLinks to an external site.
Response #1
1. Is the experience of shopping online more prone to cognitive bias than the in-store experience?
Yes, shopping online can be more prone to cognitive biases than shopping in a store. Online, companies use things like dynamic pricing and personalized discounts, which exploit biases like anchoring, framing, and loss aversion. In a physical store, you might be less influenced by these digital tactics and more by the actual product and your immediate needs.
2. In practice, can a rational shopper protect themselves from deceit when a store displays an "original price" and then a "sales price"? How much comparison price shopping is realistic given the relative price differences (between the real "original price" and the displayed "original price")
A rational shopper can protect themselves from deceit by doing their research and not fully trusting "original" and "sales" prices. However, it's hard to compare prices effectively if the "original price" is fake. Comparison shopping is realistic to a point, but the effort needed to uncover the real prices might not be worth the small savings, especially if the differences are subtle.
Response #2
The Consumer Reports investigation shows a clear example of how anchoring, framing, and loss aversion can be used to manipulate consumer behavior. The "original prices" framing is evidently geared at persuading clients towards loss aversion and portrays heavy reliance on system 1 thinking. While these techniques prove lucrative for companies, there is a real question as to whether they constitute ethical practices. Amazon's "fictitious pricing" further points to the appearance of this being a widely adopted strategy in the digital market space. Importantly, is the lesson from Thinking, Fast and Slow that, even informed consumers may not always escape manipulative techniques once these tactics are a part of an organization's marketing architecture.
In practice, can a rational shopper protect themselves from deceit when a store displays an "original price" and then a "sales price"? How much comparison price shopping is realistic given the relative price differences (between the real "original price" and the displayed "original price")?
Although an ideal customer may be able to partially protect himself, complete self-protection is hard to achieve. Indeed, a customer should simply ignore the so-called "original price" and pay attention to the final price while comparing prices at various stores. This task, however, is hardly feasible since the vast majority of customers experience numerous time restrictions and have limited access to information per time. As a result, the price indicated remains an effective anchor point, which is likely to affect consumer judgment despite their conscious attempts to avoid such influence. In terms of comparisons, there are also serious limitations here since only partial comparisons are realistic. Namely, consumers are likely to compare only several products or the final price within a basket in several online stores, two to three maybe. Checking prices for all items in each store would simply take too much time, especially if price differences per product were relatively insignificant. Thus, rational self-protection has its limits, and while one may reduce deception, they cannot avoid it altogether.