Do airlines raise prices when you search the same flight again?

What investigators and researchers actually found

The question has been tested directly, several times, by people with no stake in the answer. The results are consistent: price changes are real, tracking-based price increases have not been demonstrated.

Two caveats belong next to that table. These studies are snapshots of practice in 2014 and 2018. And “no evidence found” is not proof that no seller anywhere does it — it means the people who looked carefully did not find it.

Why the price really changed between two searches

• The cheaper fare class sold out or closed. Airlines release seats in booking classes — traditionally up to 26 — and close the lower ones as the flight fills or as departure approaches. Your second search sees the next class up. See fare class.
• Time moved. The French investigation found prices varying with the number of seats left and the date of purchase. A day closer to departure is a different pricing context.
• The first price was cached. Comparison and metasearch sites often display a recently stored price; the live price appears only when you click through.
• You arrived through a different channel. The same investigation found a site offering a lower ticket price but higher booking fees to visitors arriving from a comparison site, leaving the total about the same.
• You changed market without noticing. A different site edition or location can mean a different point of sale and a different fare set — see why flight prices differ by country.

What personalisation does exist

The 2018 Commission study found personalisation mostly in what you are shown first, not in what you pay: 92% of the airline ticket sites it tested personalised the ranking of offers, typically based on how the shopper reached the site or on past browsing.

Explicit segments are real and legal: members-only or logged-in prices, loyalty fares, student or corporate fares. Researchers at Northeastern documented members-only hotel prices for logged-in users on two travel sites. That is price discrimination by segment, but it is visible and opt-in rather than covert tracking.

Airline pricing is also becoming more granular. IATA describes dynamic pricing as using contextual information at the time of shopping — length of stay, time before travel, competition, remaining capacity — without necessarily using personal information, and states that anonymous shopping is always possible in its Dynamic Offers model.

Does clearing cookies or using incognito help?

Not for the fare, according to the evidence above. If a cheaper fare disappeared, it is far more likely that the booking class closed than that a cookie raised it. Searching again in a private window will usually show the same higher price, because the inventory changed for everyone.

Clearing cookies can change other things: the order of results, a cached price on a comparison site, or a pre-selected currency and market. That is why people sometimes see a different number and attribute it to tracking.

How to test it properly

If you want to check a specific seller yourself, design it the way the researchers did. The difficult part is noise: prices move on their own, so a single before-and-after pair proves nothing.

• Run a control and a treatment at the same moment: one clean session and one with search history, same route, same date, same market, within seconds of each other.
• Keep the market identical. Both requests must come from the same country, so a location change does not masquerade as personalisation.
• Repeat across many routes and days and compare the share of pairs that differ, not individual cases.
• Record the full breakdown — fare, taxes, fees, fare class — so a fee change is not read as a fare change.

The tooling for this is the same as for a flight price tracker: synchronised requests from a fixed market and timestamped storage. Collecting fares at scale for alerts is covered on the fare alerts page.