Abandoned Cart Discounts: Do They Work
Two practices are frequently confused. One varies the listed price by individual, the other varies which discounts an individual is offered. They differ in prevalence and in acceptability.
The list price is usually the same for everyone
Displaying different base prices to different visitors is technically possible but uncommon, largely because the reaction when it is discovered is severe.
Prices do vary by location, device type, or time, and these are frequently mistaken for individual pricing when they are segment-level rules.
The distinction matters because a segment rule applies to everyone matching a condition, while individual pricing responds to a specific person's history.
Targeted discounts are the normal practice
Offering one customer a code and another nothing achieves a similar outcome to differential pricing while leaving the published price untouched.
Because the discount is presented as a reward rather than as a price, it is generally received positively even by customers aware that others receive different offers.
This asymmetry in how the two are perceived is the main reason the industry standardised on discount targeting rather than on price variation.
The data required is already collected
Targeting works from purchase history, browsing behaviour and response to previous offers, all of which accumulate through ordinary account use.
The most valuable signal is usually what a customer has not bought, since that identifies where an incentive could change behaviour rather than subsidise it.
Customers with sparse histories receive generic offers, which is why a new account and a long-standing one see materially different promotional treatment.
Legal boundaries constrain both
Pricing that varies with characteristics protected under discrimination law is prohibited, whether the variation was intended or emerged from a model.
Data protection rules in several jurisdictions also require disclosure where a price has been personalised through automated processing, which raises the compliance cost.
These constraints apply to outcomes rather than intentions, so a system that produces such variation indirectly is not exempt from them.
What a shopper can observe
Comparing a price across a fresh session, a different device and a different account will usually show it unchanged, since the base price is rarely personalised.
The variation appears instead in which offers, codes and reduced prices are presented, and that is far harder to observe because there is nothing to compare against.
The practical consequence is that the price of an item is increasingly less informative than the set of offers a particular account happens to be shown.