Per-player pricing and the fairness line players punish

In late 2025 Sony began quietly A/B testing per-user prices on the PlayStation Store, and by 2026 the experiment had grown to more than 190 games across 70-plus regions, including the US. The same game could cost different players different amounts. When players noticed, the reaction was not gratitude for a personalised deal. It was suspicion, and the question every studio should sit with: am I being charged more than the person next to me for the identical thing?

That reaction is the whole story of per-player pricing. The mechanics work — a puzzle game reported a 19% revenue increase after moving from static to dynamic pricing with no measurable hit to retention, according to industry case-study reporting. But the same mechanic that lifts revenue can detonate trust, and the difference between the two outcomes is not the technology. It is where you draw the line.

What "dynamic pricing" actually covers

Dynamic pricing is a broad term, and lumping everything under it is the first mistake. It spans at least four very different practices that players experience in completely different ways.

The first is regional pricing — charging less in markets with lower purchasing power. Players accept this; it is visibly fair, since a price tuned to local incomes is not a penalty. The second is offer timing — showing a discount at a moment when it is genuinely useful, like a relevant bundle right after a player gets stuck. The third is composition — assembling a bundle suited to where a player is in the game, so an early player sees a starter pack and a veteran sees something strategic. The fourth, and the dangerous one, is same-item price personalisation: charging a specific identified player a higher price for the exact same SKU because the system has decided they will pay it.

The first three are about matching the offer to the context. The fourth is about extracting more from a person because of who they are. Players tolerate the first three and despise the fourth, and they can usually tell the difference even when the studio insists it is all just "dynamic pricing."

The line, stated plainly

The line is this: change what you offer, not what you charge an individual for the identical thing.

You can vary the bundle. You can vary the timing. You can vary the price by region and purchasing power. What you cannot do, without inviting a backlash, is show two players the same item at two different prices based on a model's read of how much each one will personally tolerate. The moment a player learns that the store charged them more than a friend for the same currency pack, the revenue you gained is dwarfed by the trust you lost — and trust, unlike a price test, does not roll back cleanly.

This matters more in games than almost anywhere else because game players talk. They share screenshots, compare receipts, and run forums dedicated to exactly this kind of forensic accounting. A pricing scheme that depends on players never comparing notes is a scheme with a short shelf life and a bad ending.

Why "it lifted revenue" is not enough

Here is the trap that makes same-item personalisation so tempting. You run the test, revenue goes up, and the number looks like proof. But a revenue increase from charging some players more can mean one of two very different things, and the top-line figure cannot tell them apart.

It might mean you found genuine incremental revenue — players who happily paid a price they considered fair for something they wanted. Or it might mean you simply shifted who pays, squeezing more from a segment now quietly resentful, while the long-term cost shows up later as churn, refund requests, and reputational damage that never appears in the same dashboard as the revenue bump.

The only way to tell the two apart is to measure a price change against a genuine baseline — a group of comparable players who did not receive the change — and ask whether the players exposed to it actually generated more value over time, not just more revenue in the moment. Without that comparison you are flying on a number that confidently tells you the wrong thing. See how it works → for how measuring against a permanent baseline group turns "revenue went up" into "this specific change caused this specific lift."

How to do per-player pricing without crossing the line

The good news is that almost all of the upside in dynamic pricing lives on the safe side of the line. You do not need same-item price discrimination to capture most of the value.

Match the offer to the moment. A player who just failed a level three times and a player cruising through content are in different states and a different bundle serves each — that is personalisation players welcome, because it is help, not extraction. Tune by region honestly, so a price reflects local purchasing power rather than a guess about an individual's wallet. Compose bundles for player stage, so the store shows a small number of relevant things rather than a flat catalogue priced by identity. And when you do test a price, test it the way an honest experiment is run: against a control group, measuring real downstream value, with the ability to revert instantly if the data turns.

The studios that get this right treat pricing as a place where transparency is a feature, not a liability. They can explain why a player saw the offer they saw, and the explanation does not embarrass anyone. That is the test. If you would be uncomfortable showing a player exactly why their price differed from someone else's, you are on the wrong side of the line, and no revenue figure makes that comfortable for long.

The decision in front of you

Per-player pricing is not going away, and the case studies are real. But the version that compounds — the version that still works in two years — is the one built on offers, timing, and honest regional tuning, validated against a baseline so you know the lift is real. The version that charges identified individuals more for the same SKU is a short-term number with a long-term bill, and players are very good at finding the receipt. Decide which one you are building before the model decides for you.