Seyrio methodology
How Seyrio observes flight prices
Seyrio watches defined flight trips over time and records observed fares. A Seyrio price observation is a fare recorded for a defined flight search at a particular time. Comparing observations can show how the fare Seyrio recorded changed; it cannot predict what a future fare will do.
What is a price observation?
Each observation belongs to a particular flight search, an observation time and the currency in which the fare was observed. It records what Seyrio saw at that time. It is not a promise that the fare will still be available, a confirmed purchase or a recommendation to buy.
Seyrio samples fare information returned by its flight-data provider for the stated search. This is not an exhaustive record of every airline, itinerary or bookable offer. The selected itinerary can differ between checks even when the search stays the same.
Seyrio checks monitored trips periodically. The number and timing of observations can vary, and a gap in a history does not mean the fare stayed the same.
What does “exact trip” mean?
A comparable price history keeps the defined search together: origin and destination airports, departure date, return date for a round trip, trip type, passenger counts, cabin, stop policy, sales market and observed currency. These assumptions define which fares are comparable; a search allowing a connection is different from a direct-only search.
Seyrio does not treat prices for different travel dates as one continuous history. Fares for materially different trips are not silently pooled, and a fare in one currency is not presented as directly comparable with a fare in another.
What do the homepage numbers mean?
24-hour price movement
Across 3,678 same-market comparisons, the later observed fare was 24.6% cheaper, 47.6% similar or 27.7% more expensive than the earlier observation around the 24-hour boundary. “Similar” means a change within ±1%.
Question answered: how often was the later observed fare cheaper, similar or more expensive in these historical comparisons?
Monitoring time
A longer Watch keeps the same exact trip under observation for longer. The 24, 72 and 168-hour plans use the same planned four-hour checks; paid plans do not check more often.
More monitoring time creates a wider observation window. It does not guarantee a lower fare or a successful price observation.
Why are these analyses different?
The 24-hour movement analysis compares a later observation with an earlier one. The separate historical monitoring-window analysis below checks whether any lower fare appeared during the entire window. They answer different questions and should not be compared as if they were the same statistic. Neither tells you the chance that your own flight will get cheaper.
A historical comparison of monitoring windows
This separate analysis used the same 8,348 starting observations across 22 curated exact-search series for all three durations. Each series followed a one-adult Economy round trip from 26 December 2026 to 4 January 2027, allowing up to one stop. It describes that selected cohort, not the chance that your flight will get cheaper.
A window counts as having a lower fare when an actual later observation is below its starting fare. Searches, currencies and source markets stay separate; no missing prices are filled in.
- 24 hours
- 5,526 / 8,348 (66.20%; rounded 66%)
- 72 hours
- 6,968 / 8,348 (83.47%; rounded 83%)
- 168 hours
- 7,572 / 8,348 (90.70%; rounded 91%)
The windows overlap and share observations. They are not 8,348 independent customer experiences. Changing the coverage rules or choosing non-overlapping starts changes the results. Curated evidence also had more frequent accepted observations than private Watches, so these percentages are not customer price-drop, alert or savings probabilities.
The earlier 66% / 86% / 92% homepage figures were not reproduced with their original sample counts. The current historical comparison instead uses this frozen same-anchor cohort, with rounded 66% / 83% / 91% values and its limitations shown beside the evidence.
Frozen analysis and coverage rules
Analysis reference: SEYRIO_MONITORING_DURATION_SUPABASE_RECALCULATION_V1. Fixed cutoff: .
The wider inventory contained 12,148 accepted observations across 80 exact-search series. The comparison above uses only curated collection evidence with matching search, provider-market and currency authority. Private Watch identifiers and contact information are not part of this public explanation.
Every starting observation must have enough later history to cover all 168 hours. Each 24, 72 and 168-hour window must contain at least 3, 9 and 21 subsequent observations respectively, with no gap longer than 12 hours, including the start and end. Incomplete windows, invalid prices, conflicting timestamps and observations after the trip’s monitoring boundary are excluded. These are analysis coverage rules, not promised check counts.
Comparison observations are strictly later than the start and include the window endpoint. The analysis uses only records observed and stored by the fixed cutoff, without combining different exact searches or currencies. The original September eligibility rules could not be recovered; this is a reproducible reconstruction, not proof of the old calculation.
This is an aggregate historical audit. It does not create a new immutable public fare snapshot or change the separate public trip-publication authority.
Any lower fare is different from a PRICE_ALERT. Customer price alerts require at least a 0.5% decrease against the relevant alert baseline, plus the other alert eligibility rules. Not every lower observation produces an email or Push notification. An observed reduction is not a purchase or a realized saving.
Historical observations, not predictions
Historical observations describe what Seyrio recorded; they do not predict what a future fare will do. A lower fare observed for one trip does not mean another trip will follow the same pattern.
Seyrio observes fare information, not whether someone bought a ticket at that fare. An observed reduction is not a realized saving. Flight fares and availability can change, so an observed fare is not a guarantee that the same fare can be booked later. Recheck before booking.
Reading a public price history
Where Seyrio presents an approved public trip history, “Observed by Seyrio” identifies the source and “Last observed” gives the time of the latest recorded fare. Observation times matter because prices can move. The history describes only the approved observations for that defined trip and currency; coverage may differ between markets.
Each statistic belongs to the dataset and period stated alongside it: latest is the last recorded fare, minimum and maximum are the lowest and highest recorded values, and median is the middle value after sorting fares (the average of the two middle values for an even count). They are observation statistics, not fares across every possible travel date.
A frozen snapshot keeps its original observation window and fares. “Latest in this snapshot” does not mean today’s price. Current public-market observations are shown separately with their own timestamps; a chart window and a longer statistical history must not be treated as the same dataset.
The Paris–New York evidence page offers 7D, 30D and All views from one fixed extended capture. The day windows include observations after the capture watermark minus that many days, through the watermark. All uses the whole reviewed dataset. Statistics and the table use only the selected subset; the original seven-day publication remains separate, unchanged evidence. Periodic checks may leave gaps: no missing fares are interpolated.
Explore the Paris–New York historical price evidence for a worked example with its exact search scope and complete observation table.
Public price evidence remains separate from a person’s private Flight Watch and its history. A public history does not include private Watch or contact information.