Hopper looks like a flight search app. Its economics say otherwise. The majority of what Hopper earns comes from selling optional add-ons — Price Freeze, Cancel For Any Reason, Price Drop protection, Flight Disruption Guarantee — attached to bookings made in its app. Understanding that is the difference between using Hopper well and being managed by it.
The prediction engine
The original product was a model that told you whether to buy now or wait, with a confidence indicator. It is still there and it is still the best free version of that idea, presented more clearly than KAYAK's or momondo's equivalents.
How good is it? Directionally useful. Fares on stable, competitive routes behave predictably enough that a model helps; fares on thin routes or around events do not. Hopper is honest enough to show a confidence level, which is more than most. Treat a "wait" recommendation as a reason not to panic-buy rather than as a promise.
Price Freeze, examined
This is the flagship. For a fee starting around $5 and rising with the fare and the freeze duration, Hopper holds the current price for a set window. If the fare rises, you pay the frozen price and Hopper absorbs the difference — up to a stated cap. If it falls, you pay the lower price.
It is a genuinely useful product for a specific situation: you are fairly sure you are going, you need a few days for a leave approval or a travelling companion to confirm, and the route is volatile.
Two caveats that the marketing does not lead with:
The cap. Coverage of the increase is limited to a stated amount. On a route that moves violently, the rise can exceed it and you pay the excess.
The fee is gone either way. If the fare does not move, you have paid for nothing, which is exactly how insurance works and exactly how people forget it works.
The rest of the add-on menu
Cancel For Any Reason refunds a percentage of your fare if you cancel for a reason no airline would accept. Read the percentage, and read the deadline.
Price Drop protection refunds the difference if the fare falls after you book, up to a cap.
Disruption Guarantee rebooks you if a flight is cancelled or badly delayed.
Each of these is a reasonable product. Bought together and stacked on a fare, they can add a material percentage to the ticket price, and the interface is designed to make each one feel like a small sensible addition rather than part of a running total. Decide which single risk you actually want to transfer, buy that one, and decline the rest.
Carrot Cash
Carrot Cash is Hopper's credit currency, earned on bookings and on certain add-ons, and spendable only inside Hopper. It accumulates faster than most travel rewards, which is the point: it is a retention mechanism rather than a rebate.
Value it accordingly. Closed-loop credit is worth less than cash because it constrains where you shop, and Hopper is not consistently the cheapest place to shop. A Carrot Cash balance that nudges you into booking a hotel in-app at a price you would not otherwise accept has cost you money.
Where the fares actually come from
Hopper is an agency. You are buying from Hopper, not from the airline, and changes and cancellations run through Hopper's app. Support is chat-based with limited escalation, which is fine when nothing goes wrong and frustrating when something does.
For a simple return ticket at a price the airline also offers, booking with the airline is the better contract. Hopper earns its place when you want one of its products — usually Price Freeze — and the fare is competitive.
What the prediction is actually modelling
Hopper's buy-or-wait signal is built on historical fare behaviour for a route, adjusted for how far out you are searching. That means it is most confident where there is most history — dense, competitive, year-round routes — and least useful where there is least.
Three situations where we would ignore it entirely:
Around a known event. A model trained on ordinary demand does not know about a festival, a conference or a school holiday shift.
On thin routes. Two flights a week between secondary cities produce fares driven by a handful of decisions, not by a market.
Inside about three weeks of departure. Fares in that window are governed by inventory management rather than by any pattern, and no consumer-facing model has visibility into it.
Where it is genuinely useful is the middle distance on a busy route: two to four months out, where the question "is this a normal price" has an answer and Hopper knows it.
Reading the add-on stack honestly
The efficient way to evaluate Hopper's protection products is to ask what specific risk you are transferring and what it costs as a percentage of the fare.
A £6 Price Freeze on a £400 fare is 1.5% to hold a decision for a few days — cheap. Cancel For Any Reason at 10–15% of the fare to recover a partial refund is expensive, and on a refundable fare it is redundant. Price Drop protection is worth having only on routes you believe are likely to fall, which is the opposite of what the prediction told you if you bought now.
Buy at most one. The interface will offer you all four.
Verdict
6.1. Hopper is well engineered and its core insight is sound: travellers will pay for certainty, and nobody else was selling it cleanly. The prediction display is the best free version available, and Price Freeze solves a real problem.
The score reflects the structure. A booking app whose revenue depends on selling you optional protection is an app whose interface will always be nudging you toward optional protection, and the stacked add-ons can quietly turn a competitive fare into an expensive one. Use the free prediction, buy at most one add-on, and check the airline's own price before you tap through.
