How prices get hidden.
A difference is a starting point. The evidence helps show what changed—and what remains unknown.
They know who’s buying.
You only see the price.
Sellers can use data about you when they set a price. You get one number, with nothing to compare it to.
What the seller may know
- Device
- iPhone, latest model
- Location
- Your approximate area
- Account
- Signed in since 2021
- Past purchases
- 14 hotel bookings
- This visit
- 3rd look at this room
- Estimated budget
- Higher than most
What you can see
IllustrationStandard king · one night
$229
per night
No other prices. No reason given. In this illustration, a laptop, signed out, was offered $189 the same minute. That difference alone does not tell us why.
Costnip adds the missing side: what other people were charged or shown for the same thing, at the same time.
Illustrative data, not a report about a real hotel or its use of personal information.
Real casePublished study · December 2025
437 shoppers compared carts.
Two weeks after they went public, the price tests ended.
Volunteers in four US cities compared the same groceries from the same store at the same time on Instacart, recording the prices they were shown.
- 74%
- of tested items appeared at multiple prices
- 23%
- largest observed difference on an item
- $1,200
- estimated potential annual impact for a family of four—not measured savings
- September 2025
Coordinated shopping tests at Target and Safeway stores.
- December 9, 2025
Consumer Reports, Groundwork Collaborative and More Perfect Union publish the results.
- December 22, 2025
Instacart announces it is ending item price tests.
You can help document differences too. Check an offer on two devices and record what stayed the same and what changed.
Run a two-device testInstacart said the tests were randomized and did not use personal data. A two-device check cannot by itself establish the cause of a price difference. Sources: Study and methodology; Instacart’s announcement.