Travel Agency Commission Rates and Margins in 2026

A cruise line advertises 15% commission.

In May 2026, ASTA published an industry brief that works a real booking through the actual arithmetic. A $2,000 fare for two passengers carries $120 of taxes and $100 of non-commissionable fees per passenger. That reduces the commissionable base to $1,560. The 15% pays $234 instead of $300.

The effective rate is 11.7%.

Nothing was negotiated badly. Nobody made a mistake. The published rate simply described a different number than the one that arrived.

Commission rate is the most quoted figure in this industry and the least useful one on its own. This post covers what the rates actually are, the four mechanisms that move them, and the arithmetic that turns a rate into money the agency keeps.

The gap between the rate and the base

Commission is paid on the commissionable amount, not the trip price. Taxes, port charges, resort fees, and various supplier-imposed fees are frequently excluded from the base the percentage applies to. If you want the mechanics, the calculator page explains what counts as the commissionable amount and how to work out what your agency retains after a host or consortium takes its share.

What that explanation cannot tell you is how large the gap has become. That is what the rest of this post is for.

The short answer is that it is larger than most agencies model, it varies by product, and in cruise it has been measured.

What travel agencies actually earn per booking

The most useful published figure comes from Host Agency Reviews (opens in a new tab), which has run an annual travel advisor survey since 2016. Its most recent editions draw on roughly 4,000 advisors across hosted, independent, employee, and franchise models.

Their measured result: advisors earned an average of $438 in commission per booking in 2023, on an average sale of $4,375. That is a blended rate of about 10%.

Blended is the operative word. The composition varies enormously by product:

That last row deserves an explanation rather than a number.

You will find pages quoting confident ranges for tour and package commission: 20% to 35% is the one that circulates most. We could not trace any of those figures to a named source with a stated method. Tour and package commission is negotiated contract by contract, and there is no public dataset. An invented range would be easier to publish and would tell you nothing.

If your agency sells packages, the only rate that matters is the one in your contracts.

Non-commissionable fares: the deduction most agencies underestimate

This is the largest and least visible deduction in cruise, and it now has hard numbers behind it.

ASTA's brief, The History, Economic Impact and Future of Non-Commissionable Fares in the Travel Industry (opens in a new tab), published 12 May 2026 and researched with Phocuswright, found that:

• NCFs reduce effective commission rates by 20% to 30% against nominal levels.
• A nominal 10% commission is realised at 6.5% to 8.5%.
• The effect is regressive by cabin. Entry-level cabins yield roughly 6% to 7%, while suites approach or exceed 9%.
• Removing NCFs produces a 20% to 40% increase in commission per booking.

Read the third point again, because it inverts an intuition. The bookings that take the most persuasion and serve the most price-sensitive clients are the ones where the fee structure takes the largest proportional bite.

This is also a moving target rather than a fixed condition. Viking, Virgin Voyages, and Explora Journeys have long operated without NCFs, and Norwegian Cruise Line and Oceania have both moved to eliminate them. ASTA maintains a current cruise line NCF policy page (opens in a new tab) that is worth checking before you assume a line's position.

For an agency with meaningful cruise volume, the difference between two lines with identical headline rates and different fee policies is not a rounding error.

How host and consortium overrides change the number

Commission rarely goes to one party. ASTA's brief includes the clearest published example of the full stack, on a $2,000 booking with a $1,600 commissionable base:

Without the non-commissionable component, the same three figures become $200, $60, and $40.

Two things follow from that table.

The first is that overrides are real revenue that per-booking arithmetic usually ignores. An agency modelling only the advisor line is understating what the booking generated across the structure it belongs to.

The second is that fee structures compress everyone at once. When the commissionable base shrinks, the host and the consortium lose proportionally alongside the advisor. That is why the pressure to eliminate NCFs has come from across the distribution chain rather than from advisors alone.

On splits, the commonly described tiers run from 70/30 through 80/20 to 90/10 in the advisor's favour, with movement between tiers tied to annual booking volume. Specific thresholds vary by host and change more often than trade coverage reflects. If a threshold matters to your model, read it off your host's own current terms rather than from a summary.

On consortia, be careful what you assume is knowable. Virtuoso reports $25 to $30 billion in annual purchasing power across more than 20,000 advisors. Signature Travel Network reports $11 billion across more than 15,000, with a $2 million preferred-supplier sales requirement for membership. None of the major consortia publish their actual override rates. Anyone quoting you a precise consortium override number is describing their own agreement, not an industry standard.

From gross commission to agency revenue

An agency's revenue on a booking is built in four layers, not one.

1. Gross commission: the commissionable base multiplied by the rate.
2. Commission retained: gross commission multiplied by your agency retention rate, plus any override the agency receives at its own level.
3. Markup revenue: the margin on net-rate product, which behaves differently from commission and is often the larger number on bespoke trips.
4. Service and planning fees: revenue that is independent of supplier arrangements entirely.

Then the deductions begin.

Direct transaction costs are the easiest to model accurately and the most commonly guessed at. Stripe publishes its rates (opens in a new tab): 1.5% + 20p for standard domestic cards, 2.5% + 20p for cards issued in the EEA, 3.15% + 20p for non-EEA cards, plus 2% where currency conversion applies, and £20 per disputed transaction.

Stack the applicable ones on a five-figure international booking taken on a foreign card in a different currency, and the processing cost is a line item worth modelling rather than absorbing.

Service fees are the layer agencies control most directly, and adoption has moved. Host Agency Reviews found that 56% of advisors now charge a fee of some kind. Among hosted advisors charging consultation fees, 65% use a flat fee with a median of $75, median variable fees run $100 to $325, and 17% charge per person with a median of $50.

Those medians are lower than the fee levels most agencies believe are standard. Whether that reflects a market rate or a market-wide reluctance is a question worth asking about your own pricing.

There is a more precise way to ask it. Rather than benchmarking your fee against a median from someone else's business, work out the fee your own booking would need. The calculator returns a break-even service fee for whatever you enter: the minimum fee that moves a booking from negative to neutral contribution given your rates, costs, and hours.

For a booking that took eleven hours and generated thin commission, that number is often uncomfortable. It is also the most defensible basis for a fee conversation you will find, because it describes the trip in front of you rather than an industry average.

The cost the commission rate does not show

Every layer above is commercial. It is negotiated, contracted, and largely outside daily control.

The layer that is inside daily control is what it costs to produce the sale.

A complex booking consumes advisor time in ways that rarely get recorded: researching and gathering options, turning supplier documents into something a client can read, producing and revising the proposal, chasing suppliers for confirmations, and coordinating everything that happens after the client approves.

Revisions deserve particular attention. A proposal that goes through four rounds is not 4/3 the work of one that goes through three. Each round re-opens decisions that were already made, and re-checking a complex itinerary is slower than building it the first time.

To convert that time into money you need a fully loaded hourly cost, and this is where most agencies go wrong.

The US Bureau of Labor Statistics (opens in a new tab) puts the median annual wage for travel agents at $48,450 as of May 2024, across 65,700 jobs. Dividing that by 2,080 hours gives roughly $23 an hour.

That number is wrong for this purpose, and wrong in a specific direction.

Loaded cost includes employer contributions, software, workspace, management overhead, training, and realistic utilisation (the item most often skipped). An advisor is not producing billable trip work for 2,080 hours a year. Once you account for admin, business development, holidays, and everything else a working week actually contains, the true cost of a productive hour can be substantially higher than the salary calculation suggests.

Using the naive figure is precisely why low-margin bookings look profitable on paper.

We are not going to quote you an hours-per-proposal figure. The numbers available publicly all come from companies selling software to reduce them, which is not a source we would rely on and not one you should either. Your own hours are the only ones that describe your agency.

The proposals that never became bookings

There is a larger version of this problem, and it is the one that does the most damage to agency margin.

Everything above assumes the hours were spent on a trip that sold. Many are not. An advisor produces several proposals for every booking that converts, and the unconverted ones consume research, supplier enquiries, and revision rounds without generating a single euro of commission.

Yet their cost has to be carried by the bookings that did convert.

An agency producing thirty proposals a month at six hours each is spending 2,160 advisor hours a year on proposal production. If three proposals go out for every one that books, the true labour cost attached to each booking is three times the hours that booking itself required.

The calculator's annual impact section models this directly: proposals per month, hours per proposal, and proposals per booking. It returns annual proposal hours, annual labour cost, and labour cost per booking including the ones that never converted.

Most agencies have never seen that last number. It is usually the one that changes the conversation.

Two bookings, same commission, different outcome

The following is a modelled illustration rather than survey data. The inputs are ours. The arithmetic is the point. The rows below are the outputs the calculator returns, so you can reproduce this with your own figures.

Two bookings. Same client price, same commissionable amount, same supplier rate, same agency retention rate.

Booking B produces more than twice the contribution and roughly six and a half times the return per advisor hour.

The difference is not commercial. Both advisors sold the same trip at the same price under the same supplier agreement and the same retention rate. One charged a service fee. The other spent seven and a half more hours producing it.

Note that markup is zero in both columns, because this is commissionable product. On net-rate product the markup layer replaces the commission layer entirely and the same analysis applies to it. The calculator models all three revenue sources separately, which matters when a single trip mixes them.

Nothing on an agency's commission report distinguishes these two bookings. Gross commission is identical. They appear the same in every system most agencies run.

If you want to test which lever matters most in your own business, the calculator's scenario analysis takes a booking you have already entered and models four changes against it: adding a service fee, reducing advisor hours, increasing markup, and increasing the commission rate. Run your worst recent booking through it. The lever that moves contribution most is rarely the one agencies spend their negotiating energy on.

How to benchmark against yourself

The obvious next question is what a healthy contribution margin looks like, and the honest answer is that there is no reliable public benchmark for it.

We looked. IBISWorld holds margin data for the sector behind a licence. The margin figures that circulate freely (you will encounter an "8.2% global profit margin" and a "$250,000 revenue per employee" in several places) trace back to automated statistics-aggregator sites that cite no primary source and cannot be verified. We are not going to repeat them.

What is publicly available from IBISWorld (opens in a new tab) is market context: US travel agency revenue of $45.3 billion in 2026, growing at a 9.4% five-year compound rate, across 65,181 businesses.

That tells you the sector is growing. It tells you nothing about whether your margin is good.

Which is fine, because the industry average was never the useful comparison. Two agencies with identical margins can have entirely different businesses underneath: one selling high-volume simple product at thin margin, another selling bespoke trips at high margin and high labour cost.

The comparisons that actually inform a decision are internal:

• Across product types. Which categories produce the best contribution per advisor hour, rather than the highest headline commission?
• Across advisors. Where the same product type varies widely between people, the variable is process, not skill.
• Across time. Whether contribution per hour is improving as the agency grows, or whether growth is being absorbed by production cost.

The spread inside your own agency will tell you more than any benchmark, and you already have the data.

How to audit your own numbers this quarter

This does not require a new system. It requires five bookings and an afternoon.

1. Pick five recent bookings across different product types: one cruise, one package, one bespoke FIT, and two others that are typical for you.
2. Establish a fully loaded hourly cost for the advisors who produced them. Include contributions, overhead, and realistic utilisation.
3. Record the actual hours, including revisions and post-sale coordination. Estimates from memory will be low. Ask the advisor to reconstruct it properly.
4. Calculate estimated contribution and contribution per advisor hour for each one.
5. Compare them.

The spread will be wider than expected. That is the normal result, and it is the useful one.

Two questions usually come out of that exercise. Which product types are quietly subsidising others, and what your break-even fee would need to be on the worst of them.

The number worth managing

Agencies negotiate hard on commission rates. It is visible work with a clear counterparty, and it matters.

But the rate is largely set by contracts, supplier policy, and structures the agency joined rather than built. Non-commissionable fares move it. Overrides move it. Tiers move it. Most of that movement happens somewhere other than the agency's desk.

The cost of producing the sale is different. It is entirely internal, it is rarely measured, and across a year it moves margin more than a percentage point of commission ever will.

Ghloria exists to make that side of the equation visible by keeping the client, the trip, the supplier information, and the proposal in one operational record, so that producing a complex trip takes less of the time that never appears on a commission report.

The first step does not need us. Take five bookings and work out what each one actually kept.