Flight Price Alerts: How They Work and When They Pay Off

A flight price alert is, in the abstract, a function that watches a number and pings you when the number drops. In practice, the gap between a useful alert and a useless one is roughly two orders of magnitude in engineering, and travelers rarely know which they're using.

This post is the under-the-hood version. What flight price alerts actually do, where the latency comes from, when alerts are worth paying for, and the situations where they reliably underperform.

The two things every alert system has to do

Every flight price alert reduces to two technical problems. Watching the fare, and delivering the alert. Both have non-obvious failure modes.

Watching the fare means polling some data source on some interval. The cheapest implementation polls a cached search engine (Google Flights, Skyscanner) every few hours. The most expensive polls airline inventory directly through a global distribution system query or a direct-connect API every few minutes. The cost differential is roughly 100x. The latency differential is roughly 100x as well, and not in the consumer's favor with the cheap version.

Delivering the alert is the part most people underestimate. An email alert that lands in the user's inbox 12 minutes after detection is, for a 90-minute mistake fare, useless. A push notification that lands in 3 seconds is the only thing that works for short-window deals. The math is straightforward: if the fare lives for X minutes and the delivery latency is Y minutes, your booking window is X minus Y. When X is small, Y dominates.

What "polling" actually means

Three layers of polling exist in flight price alert services, in increasing order of usefulness.

Cached web scraping. The service queries a search engine result page, parses the HTML, stores the price, and re-queries on some schedule. This is what most free flight price alert services use. Cheap, easy, and lagging the underlying truth by hours.

GDS query. Amadeus, Sabre, and Travelport sell access to live fare and availability data through their global distribution systems. Each query costs roughly $0.001 to $0.05 depending on volume and contract. A service can poll a route every 5 to 15 minutes economically, which means a mistake fare that lives 60 minutes will be caught with reasonable probability.

Direct-connect API. Some airlines (Ryanair, easyJet, increasingly Lufthansa Group via NDC) expose direct APIs that bypass the GDS. Polling these is faster, cheaper at scale, and sometimes returns fares that don't appear in the GDS at all. Most fare drop notifications worth getting on Ryanair come from direct API polling.

A serious flight alert service runs all three, plus a fare-class diff layer that compares snapshots and flags inventory transitions, not just price deltas. The reason is that a price drop from $612 to $549 might just be a competitor matching, but a sudden Z-class opening on a route that didn't have one yesterday is a different category of signal entirely.

Push versus email, and why latency is everything

A 2024 internal study of mistake fare honoring on Twitter, by the deal aggregator The Flight Deal, suggested that a fare visible for less than 2 hours had a roughly 30% honor rate when reported on by deal sites, versus 78% for fares that lived more than 6 hours. The reason isn't the fare itself, it's that ticketing queues fill up. By hour 3 the airline is processing 200 tickets per hour on a single mistake. By hour 1 it's processing maybe 10, and your odds of getting through ticketing before the fare is pulled are much higher.

This is why push notifications dominate email for time-sensitive alerts. Email service providers introduce queueing delays. Most consumer email arrives 30 seconds to 4 minutes after send. Push notifications via APNs or FCM arrive in 1 to 5 seconds. Across thousands of alerts, that's the difference between catching the fare and reading about it.

Most free flight alert services (Google Flights, Hopper for general fares, Skyscanner) deliver by email only. Most paid services that specialize in mistake fares and short-window drops deliver by push first, email second.

When flight price alerts pay off

The honest version: alerts are not always worth it. Here's where they actually help.

Long-haul international, premium cabins, and shoulder season. This is where price variance is highest in absolute terms. A business class JFK-Tokyo round-trip moving from $7,400 to $2,400 is a $5,000 swing. Catching that one alert pays for ten years of a flight alert subscription. ARC data shows that long-haul J class has the highest intraday volatility of any cabin-route combination.

Off-peak holiday weeks. Christmas-week fares to most destinations are basically fixed at peak. But the week between New Year and Martin Luther King Jr. Day in the US, for instance, has unpredictable inventory drops on JFK-LHR and JFK-CDG. Alerts catch these where casual checking misses them.

Volatile routes you'd fly if the price was right. Vague "I want to go somewhere warm" travelers benefit from broad alert subscriptions. Specific "I need Boston to Lisbon in March" travelers benefit from per-route watches.

Mistake fares, full stop. These exist for hours. Without an alert service, your probability of catching one is near zero.

When alerts don't pay off

Peak summer Europe. JFK or LAX to LHR, CDG, FCO, FRA in July and August. The fares are set, the inventory is full, the discounts don't come. Setting an alert for July JFK-LHR economy is mostly setting an alert for $40 of variance.

Tight-date business trips. If you have to fly Tuesday and return Thursday on a specific week, you're paying business-traveler pricing regardless. Alerts won't save you.

Cheap short-haul intra-Europe. Ryanair, Wizz, easyJet trend up almost monotonically from filing date. Book early. Alerts on these routes mostly just confirm that you waited too long.

How Flyozo specifically beats free alerts

The free benchmark is Google Flights price alerts. They poll once a day, deliver by email, and trigger on a relative percentage drop. For 90% of casual travelers, that's fine. For the 10% who care about catching genuine deals on long-haul, premium, or mistake fares, the gap matters.

Flyozo polls airline inventory continuously (not daily), maintains a curated mistake-fare watcher staffed by humans who validate suspicious drops before they go out, and delivers by mobile push within seconds of detection. It also surfaces hidden routes (fifth-freedom carriers, multi-stop savings, fuel-dump itineraries) that don't appear in standard searches.

For travelers who fly more than once a year and care about either cabin or price, the math is simple. The infrastructure to do this manually doesn't exist on the consumer side. The infrastructure to do it commercially does, and once it's running you get the alerts whether you're paying attention or not.

We built Flyozo because we got tired of doing this by hand. It does, basically, all of the above for $24 a year, which is roughly the price of a single in-flight Wi-Fi pass on a flight you'll catch with one of the alerts.