Ancillary revenue in travel

Almost every commercial team in travel reports a number for ancillary revenue, and very few of them are counting the same things. That is not pedantry. The definition decides whether the line is genuinely growing or whether it has quietly absorbed something that used to be counted elsewhere.

What counts, and what does not

A thing is ancillary revenue if it passes three tests, and it has to pass all three.

It is optional. The traveller can decline it and still take the trip they booked. This is the test that removes taxes, carrier charges, and anything else that arrives whether or not it was chosen.

It is priced outside the fare. It has its own price, visible as its own line, decided after the fare was decided. An item folded into a branded fare is not ancillary revenue in the strict sense. It is fare revenue with a better description.

It attaches to a trip that already exists. Ancillary demand is created by a booking. Nobody buys a checked bag speculatively.

Unbundling is what created the category

For most of the history of commercial aviation the fare bought everything: the flight, a checked bag, a seat assignment, and something to eat. Low-cost carriers worked out that stripping those items out and pricing them individually did two useful things at once.

The first is obvious. The headline fare falls, and on a shelf sorted by price the lowest number wins the click. The second is less obvious and matters more. Items that had been free became products with their own demand curves, their own price sensitivity, and their own margin. A traveller who would never pay twenty dollars more for a fare will pay twenty dollars for a specific seat, because the two decisions are not made in the same frame of mind.

Once the comparison shelf rewards the lowest advertised number, every seller that keeps the bundle is punished for keeping it. That is the mechanism that spread unbundling across the industry, including to carriers who disliked it. The traveller did not ask for this. The sort order did.

Four things a traveller is actually buying

Category lists are usually written as an inventory of products. It is more useful to group them by what the traveller thinks they are buying, because that is what decides when an offer lands and why it fails.

Comfort is the largest group by volume, because everybody has a view about where they sit. Capacity carries the highest intent, because a traveller thinking about bags is usually already packing. Certainty sells against a feeling rather than a need, which is why it lands hardest just after a large non-refundable amount has left an account. Convenience is the odd one out: it depends on the itinerary far more than on the person, so it is the group where relevance logic earns its keep. The catalogue goes through the individual categories in more detail.

Why extras became the profit engine

Three properties, working together. None of them is about travellers becoming more willing to spend.

The marginal cost is close to nothing. A seat assignment costs the operator the seat it was going to fly anyway. A bag has real handling cost, but it is a fraction of what the traveller pays for it. Revenue that arrives with a small variable cost lands almost whole at the bottom of the page.

The fare is compared and the extra is not. A traveller with four tabs open knows the market price of the flight to within a few dollars. The same traveller has no reference price at all for a specific seat in row 12. Price discovery is what makes fares brutal, and it barely operates on the extras.

The demand arrives later. Most of the interest in extras appears after the booking rather than during it, which is exactly why so much of it is collected by whoever owns the page the traveller returns to.

Arithmetic you can check yourself is worth more than an industry total. The chart below is a worked example on round numbers, not a measurement of anything. Replace every input with one you believe.

Two things fall out of that shape, and both survive changing the inputs. The cheapest item earns the most in total, because attach beats price. And the four lines together are worth considerably more than the best one on its own, which is the argument for breadth rather than for hunting a single hero product.

An airline and an intermediary are not in the same business

One phrase covers two commercial situations that look alike on a slide and behave nothing alike in practice.

An airline owns the inventory. It sets the price, keeps the whole amount, files that price into the systems that distribute it, and sees the traveller again on its own manage-booking page whenever the trip needs attention. Its constraint is willingness to pay.

An intermediary sells someone else's product. The price is set elsewhere, the economics are a distribution margin rather than the full ticket, and the relationship with the traveller ends at the confirmation email unless something is deliberately built to continue it. Its constraint is not willingness to pay at all. It is access to the moment when the traveller is ready.

That distinction sits underneath almost every argument about ancillary strategy between a carrier and its distribution partners. The two sides are optimising different bottlenecks.

Where the revenue leaks for an agency

Four leaks, in roughly the order they cost money.

The handoff. You confirm the booking and send a reference number. That number is a key to the airline's own site, and the airline has spent years making that site good at selling. You paid for the click, then handed over the second sale along with the receipt.

The blank page. Most agencies already have a confirmation page, an itinerary view, and a reminder email. In most cases those surfaces carry information and nothing else. They are the highest intent real estate in the business, in use as a filing cabinet.

The unreal offer. An offer showing a generic cabin diagram, or a bag price nobody filed, does worse than no offer at all. The traveller notices, goes to the airline to check, and buys there instead. Content sourced for that specific itinerary, through filed ATPCO content or NDC, is the difference between merchandising and guessing.

Servicing counted as pure cost. Every seat request handled by an agent is a sale that happened with no revenue attached to it. A paid support tier turns part of that cost line into a product. The OTA view works through what that changes for an agency in practice.

Sizing it without inventing a number

1. 1

   Start with bookings you already confirm

   Not sessions and not searches. Confirmed air bookings in a month, because a post-booking offer only exists once there is a booking to attach it to.

2. 2

   Apply an attach assumption you believe

   Pick a fraction you would defend to your own finance team. One traveller in twenty is a conservative starting point for a single category. Write down why you chose it, because you will revisit it once real numbers arrive.

3. 3

   Use prices you can verify

   Take a real bag price and a real seat price from your own top routes rather than an average out of an industry report. Averages hide the fact that price varies more between carriers than it does between travellers.

4. 4

   Take the deductions honestly

   Some of that revenue is not incremental, some travellers would have bought anyway, and some sales bring a support contact with them. A model that survives those deductions is one you can put in a board pack.

The result is not a forecast. It is a floor, built out of inputs you own, and it is far more persuasive internally than any figure that arrives with a vendor logo attached to it.

The window where most of it is won

Nearly everything above converges on the same place. Demand for the extras turns up after the fare decision is finished, the traveller is easier to persuade once the trip is real, and whoever owns the page they return to collects the money. That window, and how to work it without wearing out your welcome, is the subject of post-booking upsell.