The window outlives the identity
A hotel guest decides in 36 days. A cruise guest can take 18 months. Safari forgets in seven. Everything downstream of that mismatch is downstream of a measurement problem nobody priced.
25 July 2026 · Note
There are two clocks running on every considered travel purchase, and almost nobody puts them on the same axis.
The first is how long the traveler takes to decide. On SiteMinder’s platform the global average hotel booking lead time is 36.4 days, inside a typical range of 20 to 60. Vacation rentals sit under 30 days as a median, with heavy seasonality: over half of July–August bookings are placed more than 30 days out, while half of January–February bookings land inside two weeks. Cruise is a different order of magnitude. Only 11% of cruise travelers book a year or more ahead, but more than 40% of North American advisors report client windows stretching to nine or eighteen months.
The second clock is how long the browser remembers them. Safari blocks third-party cookies without exception. It deletes all script-writable storage, meaning JavaScript cookies, LocalStorage, IndexedDB and service-worker caches, after seven days of browser use without interaction with the site. And when the arrival URL carries click identifiers, which is the shape of every tracked ad click, it caps JavaScript-set cookies at twenty-four hours.
So the identity horizon is somewhere between one day and seven. The window is between 30 days and 550.
The correction that keeps not propagating
A lot of the guidance still circulating describes a flat seven-day expiry on JavaScript cookies. That was ITP 2.1, in 2019. The current rule is interaction-conditioned rather than time-capped, which sounds like it helps and does not: a shopper who researches in February and deposits in August will go seven days without a visit many times over.
Chrome’s April 2025 reversal on third-party cookie deprecation gets read as a reprieve. It is not one. Third-party cookies never solved cross-device identity in the first place, and the display retargeting they enable decays over exactly the interval that matters. The randomized field experiment puts the four-week return-rate lift at 14.6%, with a third of the first week’s effect arriving on day one and effectiveness falling from there.
What this actually costs
Not the retargeting spend. The interesting loss is upstream of that.
The ad that started the trip was generated from an inventory row. It knew the ship, the itinerary, the departure date, the lead fare, the cabin class, and what the creative depicted. Then the click resolves, and the landing surface knows that a click occurred.
Every operator I have worked with treats that as a tracking problem. It is a merchandising problem. The most specific thing you will ever know about a guest’s intent is the unit they engaged with, and it is thrown away at the exact moment it becomes actionable.
Cut from: Carrying creative semantics across the travel consideration window
