Searching in incognito mode does not lower your airfare, but the belief survives because the real causes of price jumps are invisible to travelers, and the industry is now building tools that could make a version of the myth true.
The myth does not survive the evidence
The belief is simple and widely shared. An airline or booking site tracks your searches, notices that you want a particular flight, and raises the price, so a private window or cleared cookies supposedly resets the fare. The tests we found do not support it. In 2016 Consumer Reports ran 372 flight searches on an incognito browser and on a regular one, and the price was exactly the same 88 percent of the time. Incognito showed a cheaper fare about 7 percent of the time and a more expensive one about 5 percent, as Gizmodo reported. Gizmodo also cites an academic study of one airline’s pricing algorithm, which found that overall bookings on a flight affect its price while one person’s interest does not. The most thorough test we found is a 2014 study by researchers at KU Leuven and Inria. Over 21 days in January and February 2014 it sent more than 130,000 queries to 25 airline websites from 66 user profiles in New York and Leuven, and reported that it did “not find any evidence for systematic price discrimination.” That study is more than a decade old and predates AI pricing, so it describes the system as it was, not as it is becoming.
What travelers see is real, but the cause is different. Airfare is sold in fare classes, each with its own price, and when the cheapest class on a flight sells out, the price moves to the next one for everyone looking at that flight. Fares also move constantly. CheapAir found that an average economy fare can change up to three times a day and 49 times in total. A traveler who checks twice and sees a higher number has an explanation that needs no cookies at all.
Who benefits from the myth
A belief this durable tends to serve someone, and this one serves a clear purpose by default. It points travelers at a variable they can control, their browser, and away from the ones they cannot: how fare classes are set, how demand is forecast, and which price the system decides to show. A traveler who blames cookies is not asking about any of that. We found no evidence that airlines or booking sites promote the belief on purpose, and this piece does not claim they do. Travel businesses do repeat it, though. One ski travel company’s advice, carried by Luxurious Magazine, told readers that clearing cookies could bring the same fare down to £300 or even £280. That is a single example, and it shows how easily the tip travels, not that anyone planned it.
The more interesting question is what the myth does to the debate ahead. If travelers have learned that price tracking is folklore, a real case of personalised pricing becomes easier to dismiss and harder to prove, because the claim has a long record of being wrong. A myth that is false today can make the right question harder to ask tomorrow. That is the argument this piece makes, and it is TDN’s reading of the evidence, not a finding of any study.
What moves your price today is where you are
Geography, not browsing history, is the variable with documented effects on the fare you see, alongside currency, taxes, availability and sales channel. Airlines file different fares by country of sale, so the same flight can cost one amount when bought in one country and another when bought in a second. Cranky Flier’s example is a fare to Gaborone that differs depending on whether it is bought in the US or in Botswana, and a travel newsletter notes that a Dubai to New York ticket can price differently depending on whether it is bought in India or in the US. The Globe and Mail has reported on Canadians paying hundreds of dollars more because of their country of origin, with the stated rationale that markets differ in demand and affordability.
Some carriers have moved against the practice. Lufthansa Group converted its steering logic to Point of Commencement worldwide, so the place of departure, not the country office ID an agency uses, decides availability and fare. A South African agency head told Travel News SA that this would level the playing field for smaller agencies that could not get around point-of-sale fares.
The same KU Leuven study found location differences for two carriers, and the authors treat them differently. United’s minimum price in Belgium was up to $67.92 higher than in the US for the first 15 days of the test and later dipped below it several times. The authors describe this as an exception to their overall finding but do not explain it or label it discrimination, and note that Belgian prices were converted to dollars at the applicable exchange rate. American Airlines’ Argentine site was consistently about 19 percent higher than its US site, which the authors attribute to an Argentine tax on credit card and holiday package purchases introduced in March 2013. Many other apparent differences came from availability changes, currency conversion, taxes and price caching. The study compared price time series in a volatile market, so it is best read as strong evidence against broad cookie-based discrimination, not proof that no individual price difference ever had another cause.
Currency conversion is a documented regional mechanism too. In Nigeria, international fares are priced in US dollars and converted to naira at the rate at which the central bank sells dollars to banks, which IATA says it simply applies. When that rate moved from ₦776.90 to ₦803.90 per dollar in July 2023, a dollar-priced fare cost roughly 3.5 percent more in naira with no change in the fare itself, according to The Condia. That is currency mechanics, not discriminatory pricing, but it shows how the place of purchase changes what a traveler pays.
The regional evidence is thin. We found no published comparison of point-of-sale fares for travelers in Africa or the Gulf, so the size of any gap for a traveler in Kigali, Lagos or Dubai is an open question that public sources do not answer.
The distinction that matters is between pricing a market and pricing an individual. Country of sale can influence the fare available, while taxes, currency conversion, availability and distribution channel can affect the final price. That is why a different country site may show a different price even when an incognito window does not. A country-specific fare is not necessarily one price offered to everyone in that country, because agency, channel and fare availability also vary, but it is set for a market and not for a named buyer. The debate it raises is already live: is it fair for a traveler in Kigali or Lagos to face a different price from one in Frankfurt for the same seat?
What is coming is pricing by who you are
Country of sale prices a market. Identity pricing would price the individual, and the plumbing for it is being built. NDC is a messaging standard that lets airlines send richer, more flexible offers to sellers. It can carry offers priced on criteria such as loyalty status, corporate affiliation and travel history at the time of booking, which the EDIFACT environment could not do, but whether any airline does so depends on how it builds its own pricing and retailing systems. Capability is not evidence of practice. ATPCO describes its continuous pricing approach as enabling fully personalised offers with minimum and maximum prices or bid pricing, while NDC provider Duffel tells buyers that personalisation is not available from airlines today and that connecting now is preparation for when it is. We found no documented case of an airline using an individual’s identity or browsing history to set a fare.
The best-documented NDC example points the other way. Lufthansa Group’s continuous pricing sets fares outside the traditional 26 buckets, and its documentation says the prices rest on general market factors such as season, travel day and time between booking and departure, not on the individual. The lower price is available only through its NDC API, and agents using a GDS do not see it. That is a distribution story as much as a pricing one, and a reminder that continuous pricing and personalised pricing are not the same thing.
The fight in the United States is about where one ends and the other begins. Delta told lawmakers in an August 2025 letter that no fare product it has used, is testing or plans to use targets customers with individualised prices based on personal data, and that its AI pricing tool, developed by Fetcherr, works from aggregated data and recommends changes in both directions. Critics respond that an airline does not need to know anything private to infer willingness to pay. Route, date, device, sales channel, loyalty status and in-session shopping behaviour can do much of the work, and airlines and regulators may argue for years over which of those inputs count as personal, as the blog View from the Wing argues. Regulators draw the line between dynamic pricing, which responds to market conditions, and surveillance pricing, which responds to characteristics of the individual.
The pressure is real but unfinished. In August 2026 Frank Pallone, the ranking member of the House Energy and Commerce Committee, expanded his surveillance pricing inquiry to eight US airlines, asking what customer data they use to set prices, whether AI or machine learning informs fares, and what role loyalty data plays. The letters, announced on 12 August with responses due on 25 August, followed his May 2026 letters to 25 corporations. These are requests for information, not a finding that any airline prices unlawfully.
The legal footing involves three separate questions. The first is New York’s Algorithmic Pricing Disclosure Act, enacted in June 2025, which requires sellers that set prices using personal data to display the notice “This price was set by an algorithm using your personal data.” A federal district court dismissed a retail trade group’s First Amendment challenge in October 2025, and the group’s appeal was pending at the Second Circuit in February 2026, and a case tracker still lists it as pending. That ruling concerned the First Amendment only. The second question is whether such laws can apply to airlines at all. One analyst argues that state laws of this kind are preempted for airlines by the Airline Deregulation Act, and we found no ruling that settles it. The third is the wider trend: Skift reported in February 2026 that at least 24 states were advancing algorithmic pricing bills. The industry is not united either: Reuters reported that American Airlines’ chief executive said that using AI to set ticket prices could hurt consumer trust.
Everything above concerns US regulation and US carriers. We found no comparable scrutiny of airline pricing in Africa or the Gulf.
The wrong variable
The incognito myth is wrong, and in TDN’s view that is the least important thing about it. Travelers argued about cookies while the inputs that do set a price, the country of sale, the fare class that sells out and increasingly the channel and loyalty tier, went unexamined. The useful debate is where the line sits between a price that reflects the market and one that reflects the buyer, and whether travelers, agencies and regulators in emerging markets will be in the room when it is drawn. Airlines that publish which inputs set a fare will be best placed to say they price the market. Those that stay silent should expect the old myth to return as a new suspicion.
