The travel technology value tree
Travel technology sells accuracy at volume. Revenue is a small fee on an enormous number of transactions, and the cost of being wrong is charged back to you months later.
01 The same tree, this industry
Where the money is made and lost here
The structure does not change: value is profit plus how well that profit becomes cash, profit is revenue minus cost, and revenue is price times quantity. What changes is which drivers sit underneath each branch, and which system holds them. If you have not read the general version, start with the Enterprise Value Tree and come back.
The unusual feature of this tree is that quality is a direct revenue line. A misinterpreted fare rule becomes a debit memo from the airline, so accuracy is not a support metric here — it sits on the price branch, next to the fee.
Every driver below names the metric it lands on. Follow one and you get its formula, the system the number lives in, and the ways it is commonly misread.
02 Price
What decides the price you actually get
Transaction and segment fees
A small amount, multiplied by a very large number. Pricing power comes from coverage and accuracy rather than from the rate card.
Average net price net_revenue ÷ unitsPriceSubscription tiers
Recurring platform revenue, where tier mix matters more than headline pricing.
Net revenue retention (starting_revenue + expansion − contraction − churn) ÷ starting_revenuePriceContent and incentive agreements
What carriers and aggregators pay for distribution, negotiated per agreement and settled late.
Price realisation net_revenue ÷ list_revenuePriceDebit memo exposure
Charged back when a fare rule is applied wrongly. A quality failure that arrives as a price deduction.
Contribution margin (net_revenue − variable_cost) ÷ net_revenuePriceAncillary and add-on distribution
Seats, bags and services sold through the same pipe at a different margin.
Attach rate orders_containing_the_attached_item ÷ eligible_ordersPriceEnterprise and API contracts
Bespoke commercial terms against consumption that has to be measured to be billed.
Price realisation net_revenue ÷ list_revenue03 Quantity
What decides how much you sell
Bookings and segments processed
The volume unit of the business, and the denominator of every unit economic in it.
Purchase frequency orders_in_period ÷ active_customersVolumeActive agencies and organisations
Who is actually transacting this month, as opposed to who has an account.
Active customers count(distinct customers with at least one purchase in the window)VolumeAPI call volume
Look-to-book ratios that drive infrastructure cost far more than bookings do.
Capacity utilisation actual_output ÷ practical_capacityVolumeContent coverage
Which carriers, fares and ancillaries you can actually sell. Coverage is the product.
Fill rate units_shipped ÷ units_orderedVolumeNew market and interline reach
Routes and combinations that were not previously sellable — growth created by capability rather than by sales.
Market share your_units (or value) ÷ total_market_unitsVolumeAgency retention
Switching costs are real but not infinite; retention follows accuracy and support.
Customer retention rate customers_retained ÷ customers_at_start04 Cost
What it takes to operate
Compute cost per transaction
Search is expensive and mostly does not convert. Cost per look, not per booking, is the real unit.
Cost to serve (delivery + service + returns + order_handling) ÷ customers or ordersCostContent acquisition
What it costs to hold and refresh fare, rule and schedule content at the required accuracy.
Cost to serve (delivery + service + returns + order_handling) ÷ customers or ordersCostDebit memos and penalties
The financial expression of a rules error, arriving months after the booking.
Returns rate units_returned ÷ units_soldCostSupport cost per agency
Concentrated in a small number of accounts, and usually correlated with the accuracy of the content they use.
Cost to serve (delivery + service + returns + order_handling) ÷ customers or ordersCostEngineering for rule change
Carriers change rules constantly; the cost of keeping up is the cost of the product.
Time to insight elapsed time from a business question being asked to a trusted answerCostInfrastructure and availability
Uptime in a business where a failed search is a lost booking somewhere else.
Mean time to repair total_downtime_hours ÷ number_of_failures05 Cash
Where the cash actually sits
Agency receivables and settlement
Money flowing through industry settlement, on cycles you do not control.
Days sales outstanding (accounts_receivable ÷ net_revenue) × days_in_periodCashDeferred subscription revenue
Billed ahead, recognised over the term — cash before revenue, if the terms are right.
Cash conversion cycle dso + dio − dpoCashCarrier and content payables
What you owe upstream, against what you collect downstream.
Days payables outstanding (accounts_payable ÷ cogs) × days_in_periodCashR&D capitalisation
Engineering spend that builds the product, phased against the revenue it will earn.
Capex intensity capital_expenditure ÷ net_revenueCashDebit memo provisions
Cash set aside for errors already made and not yet charged.
Forecast accuracy 1 − ( Σ|actual − forecast| ÷ Σactual )CashReturn on capital
A platform business judged on what the engineering investment actually earns.
Return on capital employed EBIT ÷ (total_assets − current_liabilities)06 Where the numbers live
The systems behind the branches
A value tree is only usable once each box maps to a system and a record. These are the six that matter most in travel technology: what each one is actually for, the records inside it the tree depends on, and which branch it feeds.
| System | What it holds | Key records | Feeds |
|---|---|---|---|
| GDS and carrier interfaces | The connections that carry availability, fares and booking messages in and out. | Availability request, Fare quote, Booking message, Carrier | Price, quantity, cost |
| Fare rules and pricing engine | The rules themselves and their interpretation — the core intellectual property and the source of both accuracy and debit memos. | Fare, Rule, Condition, Interpretation, Version | Price, cost |
| Booking / PNR store | What was actually sold: itineraries, passengers, ancillaries and every change afterwards. | PNR, Segment, Passenger, Ancillary, Change | Price, quantity |
| Billing and settlement | Fees, invoices, incentives and the industry settlement flows behind them. | Fee, Invoice, Settlement file, Incentive | Price, cash |
| Support and CRM | Agencies, cases and the issues that reveal where content or rules are failing. | Organisation, User, Case, SLA | Cost, quantity |
| Platform monitoring | Latency, availability and look-to-book by consumer — the operational cost driver of the whole business. | Request log, Latency sample, Error, Consumer | Cost |
Almost every hard question in travel technology needs two of these joined. That join — not the calculation — is the work.
07 Where it leaks
Value lost between two systems
These are the losses that no single system can see, because the evidence is split across two of them. Each one is a real number that stays invisible until the join exists — which is why the tree is an integration exercise before it is an analysis one.
| Where value leaks | Why it happens | The join that finds it |
|---|---|---|
| Debit memos never traced to the rule | A memo arrives months after the booking, is provisioned centrally, and the rule interpretation that produced it is never corrected — so the same error keeps billing. | Join the debit memo to the PNR, the fare rule version and its interpretation |
| Support cost against account revenue | A handful of agencies generate most of the tickets while revenue is reported per contract, so unprofitable relationships look healthy. | Join support cases and handling time to the account's transaction revenue |
| Look-to-book invisible per consumer | Infrastructure cost is driven by searches; billing is driven by bookings. Without joining them, the most expensive consumers are unidentifiable. | Join API request logs by consumer to the bookings they produced |
| Content coverage gaps found by customers | A missing fare or carrier shows up as a support case rather than as a monitored gap, so the loss is a lost booking you never see. | Join content inventory to search requests that returned nothing sellable |
| Incentives earned but not evidenced | Carrier and aggregator agreements pay on volumes you must prove, from data held on both sides of the connection. | Join settlement files to your own booking record, per agreement period |
08 Worked example
More volume, thinner unit economics
Transactions grew strongly and contribution per transaction fell. The tree separates it: search volume grew faster than bookings so compute cost per booking rose, debit memos from one rule family were provisioned and never fixed, and support concentrated in a small group of agencies whose fee revenue did not cover the handling.
| Component | Effect | What sits behind it |
|---|---|---|
| Transaction growth | +€2.6m | Segments processed up across the network |
| Compute per booking | −€0.9m | Look-to-book deteriorated |
| Debit memos | −€0.7m | One rule family, repeated |
| Support concentration | −€0.4m | Handling cost above fee revenue |
| Ancillary distribution | +€0.5m | The branch that improved |
| Net | +€1.1m | Growth, at a materially thinner margin |
Illustrative figures, shown to demonstrate the split. The point is the shape of the walk, not the numbers — on your own data the same bridge is built from your ledger.
09 Diagnostics
Six questions to ask in travel technology
Ask them of your own team before anyone asks them of you. In most organisations at least two of these cannot be answered without a manual exercise, and those two are the plan.
- Can you trace a debit memo back to the fare rule version that caused it?
- What is your look-to-book by consumer, and what does the most expensive one cost?
- Which agencies cost more to support than they generate?
- How long does a carrier rule change take to reach production, and how do you know it is right?
- Are your incentive claims built from your own booking data or from the carrier's?
- What proportion of searches return nothing sellable, and why?
10 The metrics behind it
Definitions for every box
11 Questions
Frequently asked
Why does accuracy belong on the price branch?
Because in this sector it is charged for. A misapplied rule becomes a debit memo, which is a deduction from revenue rather than an operational cost — so quality has a direct, measurable price.
What makes fare rules so hard to govern?
They are versioned, conditional, carrier-specific and change constantly. The governance problem is exactly a data one: knowing which interpretation was applied to which booking, and being able to show it later.
Is look-to-book a cost metric or a product metric?
Both. It drives infrastructure cost directly, and it reflects how well search is matching intent. Reporting it per consumer usually reveals a small number of integrations behaving very differently from the rest.
How does agentic AI fit this sector?
It needs exactly what this tree needs: governed, versioned content and the lineage to show why an answer was given. An agent quoting a fare it cannot justify is a debit memo waiting to happen.
See this tree on your own data
Connect the systems above, define each box once, and the tree stops being a slide.