The Psychology of Transport Pricing: What Airlines Taught Us

For about fifty years, airlines have been the most aggressive applied laboratory for pricing psychology in any consumer industry. Yield management, fare buckets, frequent-flyer points, last-minute discounting, advance-purchase premiums, fare fences, anchor pricing — the language of modern revenue management was largely written in flight-operations centres between the late 1970s and the mid-1990s, and refined every decade since.

Most transport and tourism operators have inherited very little of that work. A ferry timetable from 1995 and a ferry timetable from 2026 will often show the same fare structure: one adult price, one child price, one concession band, the same number on every sailing year-round. Coach operators do the same. Some cruise operators experimented with dynamic pricing during the 2010s and rolled it back. Most small and mid-sized operators run pricing as if airlines never happened.

There are good reasons not to copy airlines wholesale. A 200-seat ferry running six sailings a day does not need a full origin-and-destination demand-elasticity model. Most of the airline-grade infrastructure is overkill for a transport operation. But about a third of what airlines learned transfers cleanly and would meaningfully change small and mid-sized transport operators' revenue mix. That third is the subject of this article.

A short history of what airlines actually figured out

The yield management revolution is usually dated to 1985, when American Airlines launched DINAMO, a real-time pricing engine that allowed the airline to vary the seat-by-seat price on a given flight based on how full it was and how far away the departure date was. The Sabre reservation system — originally built jointly with IBM in the early 1960s as a flight-inventory system — had by then evolved into the data backbone that made airline yield management possible. By 1987, People Express, the famously low-cost airline that had pioneered cheap walk-up seats, had been pushed into bankruptcy. The post-mortem attributed the failure to yield management at the major carriers — American, in particular — selectively undercutting People Express on a flight-by-flight, seat-by-seat basis, while keeping prices high on the seats People Express did not want. The crew at People Express had a single price; American had a hundred.

The lesson, told at the time as a business-school morality tale, was that pricing was no longer a single number per route. It was a function of demand at the time of sale.

The research underneath the operational shift was substantial. Robert Cross's 1997 book Revenue Management: Hard-Core Tactics for Market Domination remains the most readable account of the period and the practices. Earlier academic work by Belobaba, Smith, and others at MIT's Flight Transportation Lab in the late 1980s and early 1990s built the mathematical foundations of fare-class allocation that the major carriers operationalised. The 1992 textbook by Talluri and van Ryzin, later expanded in 2004 as The Theory and Practice of Revenue Management, became the field's reference text.

The practices that emerged in airline operations were not just mathematical. They were behavioural. Customers buying advance were a different behavioural segment from customers buying last minute — different price elasticity, different willingness to pay, different sensitivity to flexibility. Airlines built fare fences (refundability, advance-purchase requirements, Saturday-night stay rules) to separate the segments so each could be charged what the segment was willing to pay. The economics literature calls this third-degree price discrimination; the operations practice called it fare classing.

The behavioural research caught up with the practice during the 2000s. Daniel Kahneman and Amos Tversky's prospect theory work from the 1970s, which by the early 2000s had been formally applied to consumer pricing decisions, provided the academic vocabulary for what airlines were already doing operationally. Anchor pricing, reference price effects, loss aversion, framing — the airlines were running these experiments on hundreds of millions of customers a year, decades before the academic consensus crystallised. Richard Thaler's 2008 book Nudge, co-authored with Cass Sunstein, brought the language to a wider audience. None of it was invented for airlines, but airlines applied it earliest and hardest.

That body of work — fifty years of operational practice combined with twenty years of formalised behavioural research — is what small and mid-sized transport operators have largely not taken from.

What transfers cleanly to small and mid-sized transport

Five things airline pricing taught the wider transport industry that small and mid-sized operators can adopt without an MIT-grade infrastructure investment.

1. Anchor pricing — the highest price visible first

The single most powerful piece of pricing psychology airlines use is also the simplest. The first price a customer sees on a flight-search results page is rarely the lowest. The highest-bucket business class, premium economy, or refundable fare is shown first, often quite prominently. The economy price that the customer ultimately books is presented as a relative bargain. The high anchor changes the reference price against which every subsequent option gets judged.

The research on this is substantial. The classic study is Tversky and Kahneman's 1974 paper on anchoring and adjustment, which showed that arbitrary numbers influence numerical judgements even when subjects know the anchors are arbitrary. A 2010 study by Wansink, Kent, and Hoch published in Marketing Letters found that high-price-anchor menu items increased average per-customer spend by 18% in restaurant settings, even when the anchor item was rarely chosen.

What this looks like in a transport operation: a ferry operator selling a standard adult seat, a premium-deck seat, and a private cabin upgrade should present the cabin upgrade visibly first, even if 90% of customers will buy the standard seat. The cabin upgrade is the anchor. The standard seat is the relative bargain.

Most operators do the reverse — lead with the cheapest option, mention the upgrade as an afterthought. The behavioural effect of the inversion is real.

2. Peak/off-peak versioning — temporal segmentation

The second airline-grade insight that transfers cleanly is that different times of day, week, and year contain different customers willing to pay different prices. A Saturday-morning sailing in peak summer and a Tuesday 2pm sailing in shoulder season are two different products, sold to two different customer mixes, even when they run the same vessel on the same route.

Airlines have built fare-class allocation that distinguishes hundreds of micro-segments. Transport operators do not need hundreds. They need three or four. A standard peak/shoulder/off-peak split applied across the timetable, with prices that reflect each band's demand profile, captures most of the available revenue uplift.

A 2023 study from the Washington State Department of Transportation, applied to their state-run ferry system, found that approximately 39% of ferry capacity goes unfilled across the year without temporal pricing variation, with the gap concentrated in shoulder periods. Peak/off-peak pricing closes a meaningful share of that gap, both by filling shoulder sailings with price-sensitive customers and by raising the price on peak sailings where demand exceeds supply anyway.

For most small and mid-sized operators, the cost of implementation is not the pricing logic itself — it is the configuration effort. A platform that lets the operator set versioned price lists by date range, day of week, and time of day handles the work. An operator running pricing in a spreadsheet does the work manually every season.

3. Early-bird incentives and last-minute discipline

Airlines learned early that the customer who books six weeks ahead is behaviourally different from the customer who books six hours ahead. The six-week customer is price-sensitive and planning. The six-hour customer is time-sensitive and committed. The optimal pricing strategy treats them differently.

In airline pricing, the early-bird gets an advance-purchase fare with restrictions: non-refundable, no changes, must-stay-Saturday rules. The last-minute customer pays the walk-up fare, often at a premium of 200-400% over the advance price. The fence between the segments is the booking date plus the restriction.

For small and mid-sized transport operators, the early-bird-and-last-minute approach transfers as a much simpler pattern:

• A modest discount (5-15%) for bookings made more than 14 days ahead, encouraging commitment from price-sensitive planners

• The standard price for the middle window

• Pricing discipline on the last 48 hours — no last-minute discount panic when the sailing has not sold out, because the last-minute customer is the least price-sensitive segment in the funnel

The third part is the one most operators get wrong. The panic-discount in the final 24 hours before a half-empty sailing trains customers to wait, which suppresses the early-bird segment that was the operator's most reliable revenue. Airlines learned this lesson in the early 1990s and stopped discounting last-minute except in tightly controlled circumstances.

A useful heuristic: if the sailing is under-sold 48 hours out, the right response is usually a promotional push to existing customer lists, not a public price drop on the booking page. The lists are warm; the discount is contained; the public price stays consistent and the early-bird customer does not learn the wrong lesson.

4. Channel-aware pricing — different prices for different sales paths

Airlines have priced differently by channel for decades. The fare available through a travel agent, the fare on the airline's own website, the fare on Expedia, and the fare available to a corporate-account holder with a negotiated rate are all different. The legitimate logic is that the channel has different cost of acquisition, different customer behaviour, and different revenue mix to the airline.

For small and mid-sized transport operators, the channel-aware pricing question usually centres on OTA commission. A booking through Viator or GetYourGuide costs the operator 20-30% in commission. A booking through the operator's own website costs the payment-processing percentage and nothing more. The same seat sold through both channels carries fundamentally different margins. The pricing should reflect that.

The honest implementation is not to undercut the OTA on the operator's own website (which often violates the OTA's terms of service). The honest implementation is to use the operator's own website to make the higher-margin product attractive: better-value bundles, loyalty-style perks, fewer restrictions, free changes, or simply a clearer customer experience. The OTA listing focuses on price-sensitive customers; the direct channel focuses on customers who value something other than the lowest absolute price.

The deeper version of this — capping how much capacity each channel can sell, so that the direct channel always has stock available even when OTAs are full — is a configuration concern more than a pricing concern. Covered in more depth in how transport operators lose revenue without realising it.

5. Bundling and the loss-aversion frame

The last airline-derived pattern that transfers cleanly is the practice of bundling complementary products into a single price and framing the unbundled version as a loss of features.

The classic example is the airline-fare unbundling cycle. In the early 2000s, low-cost carriers unbundled the airline product — bag fees, seat fees, food fees, priority-boarding fees. The unbundled fare was lower than the legacy carrier's all-included fare. By the mid-2010s, the legacy carriers had unbundled too, and a counter-trend emerged: the re-bundling of fares into branded fare types ("Main Cabin Plus", "Economy Flex") that included the previously unbundled extras at a premium over the bare bones. The branded bundle, framed as a small premium over the bare fare, was the most-chosen option in many airlines' ticket flows.

The behavioural mechanism is loss aversion, well-documented in Kahneman and Tversky's prospect theory work. Customers offered a bundle priced as a premium over the bare-bones option experience the bundle's exclusion of features as a loss; the premium feels worth paying to avoid the loss. The same features sold as à-la-carte upgrades from the bare fare feel like discretionary upsells.

For transport and tourism operators, the bundle pattern transfers directly. A ferry-plus-activity-plus-meal bundle priced at a small premium over the ferry-alone fare tends to outperform the same components sold as separate add-ons on the same checkout page. The configuration is identical; the framing is not.

This is also where multi-modal platforms with proper package-builder support earn their keep over à-la-carte checkout flows. The bundle has to be presented as one product, with one price, before the loss-aversion frame works. Stitching the same components together as a sequence of add-ons after the customer has already chosen the base product loses most of the effect.

What does not transfer — when to stop copying airlines

The reason most transport operators run static pricing is that they have looked at airline pricing and concluded, correctly, that the full airline approach is far too much. The honest article needs to be clear about what to leave out.

Full demand-elasticity modelling. Airlines run continuous, real-time recalculation of seat prices using inventory models calibrated against the full historical dataset of comparable flights. A 200-seat ferry running six sailings a day does not generate enough demand data to support a comparable model. The mathematical infrastructure that makes airline yield management work needs a high transaction volume that small operators do not have.

Origin-and-destination network pricing. Airlines optimise across connecting itineraries and route networks. A point-to-point ferry, coach, or activity operator has no equivalent complexity. The math is simpler and the gain from network optimisation is small.

Hundreds of fare classes. A modern airline reservation system carries dozens of price points per flight, each with its own restrictions, refundability, and channel availability. A ferry or coach operator with three to five fare categories captures most of the available behavioural segmentation. More than that creates customer confusion and operator overhead with no commensurate revenue uplift.

Last-minute aggressive discounting. Airlines do this carefully and selectively. Most transport operators trying to copy it end up training their customer base to wait, suppressing the earlybird segment. The simpler model — early-bird discount and a steady walk-up price — is usually a better fit.

The full airline model was built for a USD 800 billion industry running at low single-digit margins on enormously high transaction volume. The lessons it generated were learned at a scale very few transport and tourism operators will ever match. The right response is not to ignore the lessons; it is to pick the five (above) that scale down and skip the rest.

Three implementable takeaways

A framework is useful only if it produces operator-side action. Three concrete steps:

1. Apply anchor pricing to the highest-margin product first. Whatever the most expensive offering in the operator's catalogue is — premium-deck seating, a private cabin, a luxury package — make it the first option visible on the booking page, even if 90% of customers buy the standard. The anchor changes how the standard price is perceived. This is the cheapest single change with the largest behavioural effect.

2. Build a peak/shoulder/off-peak versioned price list and apply it everywhere. Three to four temporal bands, applied to the full timetable, run for a full season. Track conversion by sailing band against the previous year's static-price baseline. The data after one season tells the operator whether to expand the band count or stay where they are.

3. Stop discounting the last 48 hours publicly. The under-sold sailing on Saturday morning is best filled by a targeted promotional email to the operator's own customer list, not a price drop on the public booking page. The list is warm; the discount is contained to people who already chose the operator; the public price stays consistent. Track the difference between targeted-list and public-discount conversion over a quarter. The numbers usually argue against the public discount.

The transport platform consequence

Almost everything described above is configuration. A versioned price list, a fare-band structure, an anchor product, a bundle, a channel rule — they sit in the platform's pricing engine, not in the operator's spreadsheet. Operators running static pricing usually do so because their booking platform makes dynamic pricing harder than the operator's available time. The platform is the bottleneck.

A platform with a real pricing engine — flat fares, consumption-based fares, attribute-based fares, versioned price lists by date range, business rules engine for promotional logic, channel rules for direct-vs-OTA — handles all five of the airline-derived patterns above without operator-side reinvention every season. JetSetGo's pricing engine is built for this kind of work; it is the part of the platform most operators discover they actually needed once they stop running pricing as a spreadsheet.

Static pricing is rarely a deliberate choice. It is usually an artefact of the platform the operator inherited. The airline industry spent fifty years working out which behavioural patterns matter and which do not. Most of that work transfers. The operators who pick up the third that transfers, and ignore the two-thirds that do not, get most of the revenue uplift without any of the airline-grade complexity.

The pricing question, in the end, is not whether to copy airlines. It is which five lessons to take. The five above are the ones that have repeatedly demonstrated revenue uplift across smaller-scale transport operations. The rest belong to the airlines that pioneered them.

Book a demo → to see what these pricing patterns look like configured for your operation.