How we price
Every number on this site comes from the same measurement, run every hour on 300 routes — 30 destinations priced from each of 10 UK airports. Here is exactly what it is, and what it is not.
A flight deal site is asking you to believe a comparison. Ours used to be worth less than it looked: on the first version of this site the "typical" price beside each fare was the fare multiplied by 1.35, which produced a tidy 26% saving on all fifteen of them and measured nothing at all. It is gone, and this page exists so the thing that replaced it can be checked.
The price we publish
Every price on this site is a return fare for one person with taxes included, departing in the next two to 180 days, for a stay of seven to ten nights on the winter sun band and ten to fourteen on long-haul. The headline figure is the cheapest we found from any of the 10 UK airports we search, which is why the departure airport changes from day to day and from destination to destination: the question we are answering is not "what does this cost from Heathrow", it is "what is the cheapest way out of Britain to this place right now".
The airport is always named next to the price, because a fare from Manchester is not an offer to someone in Bristol — and the gap between the two is routinely over a hundred pounds. Every airport that returned a fare inside our limits is listed on that destination's page, and the whole grid is at the airport comparison.
One stop, at most
A cheaper itinerary can nearly always be found by adding a second connection, and we do not publish those. Every price on this site — the headline, and every airport figure under it — is a fare of one stop or none, inside twelve hours' travelling on the winter sun band and eighteen on long-haul. It is the cheapest fare that meets that description, not the cheapest fare.
So an airport can be missing from a destination's list while flights from it plainly exist, and a destination can drop off the board for a day. That is the rule working: the alternative is a headline price attached to a twenty-two hour journey through two airports you did not choose, which is a worse holiday sold as a better deal.
Where a longer routing saves a genuinely large amount we show it separately, as a second fare, with the hours and the wait written next to it. It is never the number in the headline.
The struck-through price
Every hour, for each route, we pull thirty live fares from the same search and take their median. That median is the number you see struck through. It is what someone searching this route today, without us, would reasonably expect to pay.
It is deliberately a conservative comparison. Those thirty are the cheapest thirty of thousands of results, so the true shelf price of the route is higher than our median and the percentage we advertise is smaller than the one we could justify. That is the direction we would rather be wrong in for a number shown to buyers.
When there is no comparison, we show none
If the median does not come out above the cheapest fare, there is no saving to report and the card carries a price with no percentage and nothing struck through. Of the 30 destinations we currently hold a fare for, 24 have a measured comparison today and 6 do not. Those 6 show a plain price. We do not manufacture a "was" figure to put a line through it.
The percentage badge has a floor as well: below twenty per cent it is not drawn at all. On the winter sun band the fares are cheap outright rather than marked down, and a six per cent badge beside a fifty-eight pound flight reads as an apology for a number that needs none.
What we know about a route, and for how long
We keep the cheapest fare seen each day, per route, for the last ninety days. That record is what any "lowest in N days" badge refers to — our own observations, over a window we state on the badge itself. It is not the airline's pricing history, which we do not have and cannot see. A route we have only been watching for a short time gets no rarity badge at all until the record is long enough for the comparison to mean anything.
What is on the page before anything loads
Every page here is built with the last fare we measured, so there are real numbers in the HTML before any script runs and a search engine sees prices rather than empty slots. As the page loads it re-checks against the live feed and replaces those numbers with the current ones — the fare, the dates, the airport and the temperature all move together.
If the feed cannot be reached — you are offline, an extension blocked it, a worker is having a bad minute — the fare from the build stays on screen rather than the page blanking itself. That number can be as old as the last deploy, which is why every fare here is a link to the live search that will confirm or contradict it before you pay anything.
What this method cannot tell you
It cannot tell you where a fare goes next. We publish no predictions and no buy-or-wait advice, because we have no basis for either.
It also compares a fare only against other fares found at the same moment. A route that is expensive across the board on the week you searched will produce a small percentage even when the whole week is a bad time to fly — the comparison is against today's market, not against the route's best season. The seasonal advice on our destination pages is editorial judgement, clearly separated from the measured numbers, and labelled as such.
Who pays for this
Booking links go to Kiwi.com and hotel links to Stay22, and we earn a commission when a booking completes. The commission is paid by the platform and the price you pay is the platform's own — clicking through here does not change it.
Commission rates play no part in what gets published. The ranking rule is the measured percentage and then the price, applied to every destination the same way, and it is the code that applies it rather than an editor. Full disclosure is in our terms.