Flight alert benchmark: TripWaffle vs Flighty vs TripIt

Scoreboard

Per alert type: how many flights each app alerted on, who was first when two or more did, and each app's median timing against the aircraft. A gate change is measured before scheduled departure, a delay before pushback, landed and baggage after touchdown.

"Flights" counts flights where at least one app sent that alert. "First" is scored only on flights where two or more apps alerted. "Skipped" is a delay TripWaffle saw and held because it was under 15 minutes, its alert threshold. "False" is a delay alert that predicted 15 minutes or more, or gave no figure, on a flight that left under 15 minutes late; it is not counted as an alert, as a first, or in any median. Each app is scored only on the flights it was tracking ("of N given"). The first median is over each app's own alerts, so on delays it can compare different flights; the "same flights" median is over the flights where every app alerted, which is the apples-to-apples figure.

Every flight

Newest first. Each cell is how far the alert landed from what the aircraft did. Green is the earliest app. A dash means that app sent no such alert. False alert means the app alerted a delay of 15 minutes or more that did not happen. An app has a row only on the flights it was tracking. The flight row carries the outcome so you can judge whether an alert was owed at all.

79 older flights are not shown: the notification record this page reads holds a fixed number of recent rows, and a flight older than the oldest row would read as one-sided.

How we measure

• One device, every app. TripWaffle, Flighty and TripIt run side by side on one device. Every delivered notification is read from the operating system's own notification log, with the delivery time the operating system recorded. No app is told it is being measured.
• The same flights. Real, scheduled flights across airlines and regions are added to every app well ahead of departure, a few each day, so the set is not picked for a result. TripIt receives them as plans in a trip, the same way a traveller would enter them.
• Judged against the aircraft, not against each other. Each alert is measured from the event it describes: a gate change from the scheduled departure, a delay from the recorded pushback, landed and baggage from the recorded touchdown. Every app's alerts sit on that one clock. Who was first is a by-product.
• Delivery time, not send time. TripWaffle's own send log is paired with what the device saw, and the device time is what counts. A push we sent that never arrived counts as no alert.
• First alert per type per flight. Flighty's first gate notification is its gate reveal about five hours out, whatever it is titled, so its second gate notification counts as the change. TripWaffle and TripIt only send a change, so the first counts. A gate alert sent after the aircraft has pushed back is not scored for any app. A baggage alert that names the belt also counts as the landed alert on every app. TripIt's arrival notice (terminal and gate on landing) is its landed alert.
• A false alert is not an alert. A delay alert counts as false when the flight left under 15 minutes late and the alert predicted 15 minutes or more, or gave no figure. The rule is the same for every app, TripWaffle included. A small alert that was right, such as 10 minutes on a flight that left 11 minutes late, stands.
• A wrong gate is not a gate alert. Where TripWaffle sent one gate alert and it named the gate the flight left from, and Flighty's scored alert named a different gate, any app's scored alert that named a different gate reads "wrong gate" and wins nothing. First goes to the earliest alert that named the right gate.
• Every flight is audited. After a flight settles, an automated audit re-reads its figures against the notification record. Where a figure needs context it adds a note under the flight. A note explains a figure and never changes one. An alert that turned out wrong gets a note too, whichever app sent it and whether or not its type is scored here: a notice about an aircraft that did not fly the flight, a gate the flight did not use.
• A flight appears once it has settled: on blocks for 90 minutes, or touched down for two hours, or cancelled with the departure 90 minutes past, or six hours past its scheduled arrival with no actual time at all.

What we do not compare

Only alerts the apps send by design are scored. These are left out, and the reason is stated so the omission is not mistaken for a loss:

• Boarding soon: TripWaffle sends one, the others do not, so there is nothing to race.
• Took off, arrives at gate in one hour, arrived at gate, flight plan filed: Flighty sends these, TripWaffle does not by design.
• First gate assignment: TripWaffle prints the gate in the pre-flight brief and pushes a first gate only close to departure. Only a change is scored.
• Terminal changes: TripWaffle tells you when your departure terminal moves after your pre-flight brief. The table does not score terminal alerts for any app.
• Storm-risk forecasts: TripWaffle warns the evening before when thunderstorms are forecast at the departure airport, and withdraws the warning if they clear. Nothing comparable exists elsewhere, so it is not scored and does not count as a delay alert.
• Delays under 15 minutes: TripWaffle does not alert them. Such a flight shows "skipped" in TripWaffle's delay cell. Another app's alert stands when it predicted under 15 minutes.
• Day-of status and check-in reminders: TripIt's "on time" morning status and its check-in note, like TripWaffle's pre-flight brief and check-in push, are briefs, not alerts about a change, and are not scored.
• Notification wording: this page publishes alert types and timings only, never the text of any app's notifications.

Known limits

• The reference times (pushback, touchdown) come from the same flight-data sources TripWaffle uses. When a source records touchdown late, every app can appear to alert "before" it; the record is late, not the alerts early.
• Alerts are attributed to flights by flight number and route. A rare mismatch would put an alert on the wrong flight; the attribution rules are reviewed in every audit round and corrected when found.
• Some airports publish no baggage belt; no app can alert one there.
• Now and then an app does not pick up a flight it was given. That flight is left out of that app's figures, not counted as a miss.
• TripIt joined the benchmark later than the other two, so it is scored on fewer flights; the "of N given" figure says how many.
• This is TripWaffle's own benchmark. The method, the exclusions and every row are published so the reading can be checked, but the rig is ours.

Looking for a Flighty alternative on Android?

Flighty is available on Apple devices but not Android. TripWaffle runs on Android and iOS and is currently free, and sends the alerts scored above: gate changes, delays, landed, baggage belt, plus a boarding-soon push, terminal-change alerts and a pre-flight brief. The table is the honest answer to how the apps compare on timing; read it before deciding.

Reuse

This page is free to read, link to and quote with attribution. Cite it as "TripWaffle flight alert benchmark, as of 2026-10-27" with a link to this URL; the figures change as flights settle.