A notification appears on your phone: flights to Barcelona are now cheaper. You open it. A four-stop itinerary with a 60-hour total journey has fallen from $480 to $465. You close it.
Something changed in the market. Nothing improved about your trip.
Price alerts solve a real problem
Flight prices move. Most travelers do not want to repeat the same search every morning to see whether the fare has changed.
Alerts are useful because they remove part of that work. You define the route or flight you are interested in. The service watches the price. When something changes, it brings you back.
That is already valuable. The problem appears when the alert watches a broader set of flights than the traveler would ever seriously consider.
A cheap flight is not necessarily a relevant flight
Suppose your original request is: “I want to fly from Yerevan to Barcelona after work on Friday and return late Sunday. One connection is acceptable. I would pay a little more for direct, but not $200 more.”
That request contains much more than two airports and two dates. It includes several preferences:
- The outbound flight should leave after work.
- The return should preserve most of Sunday.
- More than one connection is not acceptable.
- A long overnight stop would damage the trip.
- Direct is preferred, but only within a reasonable price difference.
- The trip is short, so time matters more than it would on a long holiday.
A route may become technically cheaper because an itinerary outside all of these conditions has dropped in price. The alert is correct. It is still not useful.
The traveler does not care whether the cheapest imaginable way to reach Barcelona became cheaper. They care whether a flight that fits the trip became better.
The preferences should remain attached to the alert
The first version of Apricot’s alert experience is deliberately simple. The traveler sets an alert for the trip. But the flights being watched should still be shaped by the information and preferences already included in the request.
- Departure window.
- Arrival time.
- Number of connections.
- Baggage requirements.
- Direct-flight preference.
- Price sensitivity.
- The other conditions that made an option relevant in the first place.
Apricot does not need to notify the traveler because any random fare on the route moved. It should bring them back when a good option — one that could genuinely fit those requirements — becomes meaningfully cheaper.
That is the essential distinction. We are not trying to build a louder price alert. We are trying to make the alert more relevant.
“Cheaper” depends on what became cheaper
A $15 drop can be meaningless. It can also be exactly what someone was waiting for.
If a four-stop itinerary falls by $15, the traveler may still never consider it. If the direct Friday-evening flight they already liked falls by the same amount, that may be enough to book.
The value of the movement depends on the quality of the option, not only the percentage change. A useful alert therefore needs to consider two things together:
- Is this flight relevant to the trip?
- Has its price changed enough to matter?
One without the other creates noise.
Alerts should reduce checking, not create more of it
An irrelevant notification has a hidden cost. You open the product. You inspect the result. You remember why the cheap option was wrong. You check whether the flights you liked changed too. You close it.
The alert was supposed to remove repeated searching. Instead, it initiated another search.
This is the difference between monitoring the market and helping with a decision. Monitoring tells you that something happened. Assistance understands whether it deserves your attention.
What we want the Apricot alert to mean
When an Apricot notification arrives, the message should not mean: “Some flight on this route became cheaper.” It should mean: “One of the flights that could fit the trip you described has become meaningfully cheaper.”
That flight might match the departure time. It might satisfy the arrival deadline. It might stay within the acceptable number of connections. It might include the baggage conditions the traveler needs. It might be the direct option that was previously just outside the budget.
The alert is still simple. The relevance comes from using the traveler’s filters and preferences to determine which flights are worth watching.
The best alert may be no alert
Many digital products are designed to create reasons to return. More notifications produce more engagement. But a travel assistant that promises to save attention should be comfortable staying quiet.
No notification may simply mean that nothing useful changed. Then, when a good flight becomes meaningfully cheaper, the alert has a clear purpose: this option is worth looking at now.
That is what we want Apricot alerts to become. Not an update whenever the market moves. A signal when the trip gets better.