Post-booking upsell

Most travel sellers treat the moment after checkout as an administrative full stop. It is closer to the beginning of a second conversation, and it is the one where the buyer has already said yes once.

Why the window behaves differently

Three things change at the instant a booking confirms, and none of them is about the traveller becoming a different person.

Commitment has happened. The trip has moved from hypothetical to real. The dates are blocked, somebody has been told, and the money is gone. Every decision after that point is about a trip that exists rather than about whether to take one.

The price anchor is gone. During checkout, every add-on is judged against a total that is still being compared with three other tabs. Twenty dollars for a seat is twenty dollars added to a number the traveller is trying to make smaller. An hour later that comparison is closed, and the same twenty dollars is judged on its own merits.

Intent rises as departure approaches. The traveller thinks about the trip more often, not less, as it gets nearer. Packing, work handovers, and airport logistics all surface needs that simply did not exist at the point of sale.

Timing is a sequence, not a moment

The instinct is to find the single best moment and put everything in it. The window is several days or several months long, and interest moves around inside it by category.

Four moments do most of the work, and each one has a different reason to exist.

Confirmation. The trip is at its most vivid and the traveller is already on your page. This is where seating decisions land, because where you sit is the first concrete thing anyone imagines about a flight.

The week after. Second thoughts arrive. Somebody realises the return is a red eye, or that they are travelling with a colleague and should be seated together. A single, quiet reminder here catches people the confirmation page missed.

The fortnight before. Packing begins, which is when baggage becomes an actual decision rather than a hypothetical one. Ground logistics and destination needs also surface here for the first time.

An event the traveller recognises. Check-in opening, a schedule change, or an equipment swap. These are messages the traveller was going to read anyway, which makes them the least intrusive place in the whole window to attach an offer.

Which categories suit which moment

Category and timing are not independent choices. Getting the pair right matters more than getting either one right on its own.

Comfort products belong early. A seat is imagined the moment the trip becomes real, and it is the decision people most want settled. Cabin upgrades sit in the same family, though they behave differently: the largest single ticket in the category, and the one most sensitive to being raised at the wrong moment.

Capacity products belong late. A traveller two days from departure with a suitcase open on the bed has a much clearer view of their baggage than they did at checkout.

Certainty products belong immediately. Refund protection speaks to a feeling that is at its strongest in the minutes after a large non-refundable amount has left an account, and it fades quickly after that.

Convenience products belong to the itinerary rather than to the calendar. A long connection is what makes a lounge worth thinking about. A late arrival into an unfamiliar city is what makes ground transport worth thinking about. The catalogue covers how each of those categories differs.

What the offer has to be

An offer in this window is judged more harshly than one in checkout, because the traveller has no reason to tolerate it. The bar is not persuasion. The bar is usefulness.

1. 1

   Show only what fits this itinerary

   The real cabin on the real aircraft, and a price the airline actually filed for that flight. A generic diagram or an invented price sends the traveller to the airline's own site to check, and they will finish the purchase there.

2. 2

   Lead with what they were already thinking about

   On a long haul that is usually the seat. On a low-cost carrier it is usually the bag. Ordering the cart by what the traveller has already half decided costs nothing and changes the first impression of the whole page.

3. 3

   Ask once per reason, not once per week

   Every message needs a reason the traveller would recognise. Tie reminders to events rather than to a schedule of your own, and stop asking about an item they have already declined twice.

4. 4

   Take one payment and confirm immediately

   One cart, one payment, one confirmation, with the booking reflecting the purchase without anyone at your end handling it. Every extra step between wanting the thing and having it is a place to lose the sale.

Useful, not a shakedown

The difference is mostly about who the mechanics serve. Five rules cover nearly all of it.

Relevance first: never show something the traveller cannot actually have on that flight. Verifiable pricing: a number they could check against the airline without feeling cheated. No invented scarcity: countdown timers and phantom seat counts buy one sale and cost the relationship. Nothing pre-ticked: an add-on the traveller did not choose is a refund request and a chargeback waiting to be filed. And an honest decline: the way to say no should be as easy to find as the way to say yes.

None of that is ethics for its own sake. Support contacts, refunds, and chargebacks are the direct cost of an offer that oversold itself, and they are usually large enough to erase the revenue that caused them.

How to measure it

Three numbers, and a discipline about the denominator.

Attach rate is the share of bookings that buy at least one item. It tells you whether what you showed was relevant, and it is the number to watch while you tune the offer set.

Revenue per booking is the honest headline, because it holds when volume moves. Attach rate can rise while revenue per booking falls, which usually means you have started leading with the cheapest thing on the page.

Support contacts per sale is the quality check nobody runs. An offer that generates a contact for every few sales is not profitable at all once the handling time is counted.

Below is the same arithmetic as a worked illustration. The inputs are round placeholders, not measurements from anyone's account.

Read all of it against a comparable period rather than against zero. Seasonality moves attach as much as merchandising does, and a launch that coincides with a holiday peak will flatter itself. If you can afford the volume, hold a slice of bookings out of the programme entirely. A holdout is the only way to separate revenue you created from revenue you moved.

Where to start

Pick one point of sale, one surface you already own, and two categories you can show accurately. Prove the number against a period you can compare with, then widen. The commercial case for the whole category is set out in ancillary revenue, and the agency version of the argument sits in the OTA view.