How flight pricing really works: evidence vs myth

by

Last reviewed July 2026

Two things are both true here, which is exactly why the folklore around flight pricing is so persistent. Personalized pricing on travel sites is real and has been documented by academic researchers. And the specific, most widely repeated version of the myth, that searching in incognito mode or clearing your cookies lowers airline ticket prices, is not supported by the evidence. Conflating the two is where most of the bad advice comes from.

What’s actually been proven: hotel and steering, not airfare cookies

A landmark 2014 Northeastern University study examined 16 major e-commerce sites, including six travel and hotel platforms, and found real, measurable personalization on several of them.1 Cheaptickets and Orbitz offered lower hotel prices to logged-in “members,” sometimes a difference of around $12 a night. Expedia and Hotels.com steered certain users toward more expensive hotel options in search results, a practice distinct from price discrimination called price steering. This is legitimate, well-documented research, and it’s specifically about hotel bookings and account status, not airfare and browser history.

When researchers looked specifically at airline ticket pricing, the picture changed. A 2014 academic study of online airline tickets, titled “Crying wolf?”, found no evidence of systematic price discrimination tied to browser fingerprints, attributing the price volatility travelers notice to airlines’ highly dynamic pricing algorithms rather than tracking.2 That finding holds up in a larger, EU-wide study: the European Commission’s 2018 mystery-shopping study of 160 e-commerce sites across eight member states found personalized ranking of search results on 92 percent of airline ticket websites, the highest rate of any sector tested. But across all four sectors it covered, including airlines, it did not find evidence of consistent, systematic personalized pricing. Where price differences did turn up, they were small, a median under 1.6 percent, and not reliably higher or lower for any particular type of visitor.3

Why the myth feels true anyway

Airline fares change constantly, independent of anything you personally do, because of seat inventory shifts, fare bucket availability, route demand, and time until departure. When a price jumps between two searches minutes apart, it’s easy to conclude your browsing caused it. The more mundane explanation is usually correct: a lower fare bucket sold out, or you searched for a different number of passengers than the last time, which can push the whole itinerary into a higher blended fare class.

What does move the price, legitimately

  • Point of sale and currency. Searching from a different country’s version of an airline site or using a VPN to change your apparent location can surface genuinely different fares, because airlines price markets differently. This is a real effect, distinct from the cookie myth, and it’s the one situation where a VPN can plausibly change what you see.
  • Multi-passenger searches. Searching for a family of four shows the price for the cheapest bucket that has four seats left together, which is often higher than the single-passenger price you saw the day before.
  • Timing and booking class. Fares generally rise as departure approaches and as cheaper booking classes sell out, not on a fixed “book six weeks ahead” schedule, despite how often that specific number gets repeated as a rule.

Practical takeaway

Skip the incognito ritual for flights specifically, it isn’t costing you anything to do, but it also isn’t earning you anything. Track a fare over several days using a tool that pulls live inventory rather than refreshing one tab repeatedly, and treat urgency messaging (“only 2 seats left at this price”) as a sales prompt rather than proof, since it’s rarely independently verifiable. For hotels, logging in or comparing as a “member” versus a guest is worth doing, since that’s the one area where the research shows real account-based pricing differences.

Sources