Dynamic ancillary pricing that prices every paid seat, bag offer and bundle per flight, inside the rules your people set
3% more seat and bag revenue per passenger, and families always seated together free.
See one case, screen by screen ↓Airline Dynamic Ancillary Pricing is dynamic ancillary pricing for airline ancillary teams: airline dynamic pricing for paid seats, bags and bundles, set per shopper type and flight, inside rules people set. Eight agents read demand, propose a price for every paid seat zone, bundle and bag offer, check it against the guardrails, re-compute it and design a test cell. The ancillary pricing manager approves, caps or rejects, and the director co-signs above $25,000.
How to choose AI for airlines: a buyer's guide to agentic AI solutions →
Updated
Ancillaries are nearly a sixth of airline revenue, with seats, bags and bundles at their core, yet most network carriers still price them from fee tables that change a few times a year instead of with dynamic ancillary pricing: one exit-row price on every Chicago–Orlando flight, whether it sells out a month early in spring break week or flies half empty on a February Tuesday. Before a peak, an analyst exports take rates by zone, compares last year in a spreadsheet and proposes a seasonal uplift for a route group; the decision sits in someone's inbox.
Nobody checks the new price against family seating, exit-row eligibility under 14 CFR 121.585, disability seating under Part 382, elite and corporate waivers or refund rules before it goes live. Price tests are rare, run without a written stop rule and read weeks later. When network planning swaps an aircraft, paid seats disappear and the airline learns about it from complaints — though a fee for a seat the passenger paid for and did not get must be refunded under the DOT refund rule, 14 CFR Part 260. One price that charges a parent to sit with a child does more harm than the uplift is worth, which is why most teams have not moved.
Estimated split for one route group's seasonal change, by hand and with the solution.
Eight agents read demand, price each seat zone, keep the rules, re-check and test; managers decide.
Corvane Airways flies out of its Chicago O'Hare hub. It is Thursday, October 15. Overnight the agents priced every future flight-date with paid seats, and Dana Whitfield, the ancillary pricing manager, has six decisions to make. Here is the first big one, screen by screen, in the working solution.
Dana's morning opens on four numbers: 41,812 flight-dates priced overnight, 100% inside the guardrails; 6 decisions for her, est. +$62,700 on the priced ones; 2 tests running; 2,940 children seated beside an adult at $0 this week, none charged. Below, the list is sorted by decide-by time — an aircraft swap on CV 1182 by 15:00, New Year's Eve exit rows where the Checker differs by 17:00, spring break seats to Florida by Friday 12:00 — and beside it, the Seat pricing agent's summary with every claim cited.
“Start with the aircraft swap on CV 1182: 14 people paid for seats they will not get, and the refund clock runs from today.”
AP-2207 covers 46 spring break flight-dates, Chicago to Florida, March 13–21, 2027. Dana opens one: CV 734 ORD–MCO, Saturday March 20, a 737 MAX 8 with 172 seats, 71 booked, 156 days out; every seat sold through paid seat selection shows the proposed price in bold with today's faded beside it, under the guardrails people set. Exit rows $79 → $95, extra legroom $69 → $84, preferred aisle and window $29 → $36, standard aisle and window $15 → $18, preferred middle held at $19: seats still to sell on this flight $2,079 → $2,329, +$250. The Guardrail keeper notes that extra legroom and preferred move more than 20%, so they post in two steps: $83 and $35 on day one, the full price the next.
“Exit rows $79 → $95. Last spring CV 734 sold its exit rows out 31 days before departure at $79; this year spring break Saturdays are 18% ahead.”
Dana switches the map to what a family with children sees. The exit rows show as not offered to a party with a child: 14 CFR 121.585 bars anyone under 15 from an exit seat, and the airline's rule keeps the whole party out of the exit rows. Each child 13 and under sits beside an accompanying adult at $0 on every fare. On CV 734 the Guardrail keeper counts 11 children in 6 families and holds all their seats at $0. It also notes that no personal data was used: prices are set per shopper type, from trip context only.
The family seating rule is locked. No agent and no pricing manager can put a price on it.
The Checker re-computed the proposal on its own, without seeing it, and lands at $93 for exit rows — agreeing within $2, confidence 0.82. The Test designer proposes T-41: 10% of shopping sessions for 21 days, assigned by session and never by person, with current prices as control. Dana asks why not $89, the price Larkspur Air shows on its 07:30. At $89 the take model expects 7.9 exit seats sold, 0.3 more than at $95, but $703 against $722 — $19 less on this flight, about $870 across the 46.
Stops if exit-row take falls under 70% of control, any family is charged to sit together, or seat-fee complaints double.
Dana would rather match the market while the test runs. She caps exit rows at $89; the Checker re-computes — CV 734 +$231, 46 flight-dates est. +$10,600, inside its $86–$97 band. The approval asks for a test cell (5%, 10% or 20% of shoppers), a reason code and her confirmation: she picks 10% and CP, capped below the proposal. Shoppers in the test cell see the new prices from Friday 06:00; everyone else keeps today's as control.
“I, Dana Whitfield, make this decision under my name. It is logged and checked after departure.”
Every proposal carries an evidence pack: the 6 passages cited, read-only copies; 9 of 9 guardrails checked with the rule text; the Seat pricing agent's workings and the Checker's independent re-computation; the shopper-type inputs used, with no personal data; the test plan with its stop rules; the sign-offs; and the audit trail. It downloads as one 14-page PDF.
CV 1182 ORD–SFO on Saturday November 21 changed from a 737 MAX 9 to a 737 MAX 8. The Fee refund watcher read the aircraft change, compared the two seat maps — 36 extra-legroom seats down to 24 — and found 14 passengers who paid $79 for extra legroom they can no longer get. It chose new seats, kept the Ramirez, Thompson and Lindqvist families together with 4 children beside an adult at $0, and prepared 14 refunds, $1,106, to the original payment under 14 CFR Part 260.
“Waiting for Dana Whitfield — nothing is refunded or moved without a person.”
The Bag and bundle agent proposed raising the first checked bag at the airport from $40 to $45 in spring break week, est. +$6,300. The Guardrail keeper blocked it: a bag fee is a published fee, the same for every shopper and shown on the first fare display, and the agents never change one. It gives Dana the route forward — send it to Distribution as a fee-change request; Distribution files changes with 24 hours' notice to every channel — or a prepaid bag offer at $35 online inside the published fee.
“The agents never change a published fee. Distribution decides and gives every channel 24 hours' notice.”
Small, safe moves should not eat her morning. Select the low-risk ones picks four, est. +$12,300: Thanksgiving standard seats ORD–DEN $15 → $12, already proven in test T-29; off-peak Tuesday preferred seats $29 → $25 on 214 flight-dates; extra legroom $49 → $55 on early business flights to LaGuardia and Reagan National; and seven more days for the Family bundle test. The Checker agrees on all four, all in rules; one sign-off with a reason code logs each.
The Outcome reviewer replays every departed flight that had a price change. The same move on Labor Day exit rows, ORD–Florida, earned $9.84 paid-seat revenue per passenger in the test cell against $8.71 in control, +13.0%, with 418 of 418 children seated at $0. Over 30 days: 14 decisions checked, 10 helped, 2 even, 2 cost, an estimated net +$186,400 against control, with test cells scaled to all shoppers. It suggests allowing a ±25% daily move on peak leisure Saturdays — and waits for Graham Pierce, because only the director changes a guardrail.
“Family seating, exit-row and refund rules are unchanged.”
The same spring break week, seen by five of the people who carry it — what pricing looked like, and what it looks like now.
Eight agents read demand, price each seat zone, keep the rules, re-check and test; managers decide.
Reads bookings, seat maps, shopping sessions in aggregate, last year's flights, the events calendar and rival seat and bag fees every night, and links each signal to the flights it touches.
Proposes a price for every paid seat zone, flight and shopper type from the take-rate model by days to departure, and writes why in plain words with each source cited.
Prices bundles and bag offers, such as prepaid bags, per route and shopper type, inside the published checked bag fees.
Checks every price and offer against every rule — floors and ceilings, the largest daily move, family seating at $0, exit rows, disability seating, refunds, waivers and the same price in every channel — and blocks a breach with the rule and the route forward.
Re-computes every proposal from the raw data without seeing it, and compares its answer with the agent's.
Proposes a test cell, control and stop rules with every proposal, assigns shoppers by session, checks each running test every morning and stops one that breaks a stop rule.
After departure, compares the test cell with control on every flight that had a price change, says what helped, what was even and what cost, and suggests what to keep.
Watches aircraft swaps and seat-map changes, finds every passenger who loses a paid seat, prepares the Part 260 refunds and reseats families together.
Ask why, cap, approve with a reason code; the director co-signs above $25,000. The agents propose; a named person decides.
Ask the agents in plain words on any screen. They answer with numbered sources, open the screen that answers it, and re-compute when you set a cap.
What should I look at first?
By time: AP-2218 refunds by 15:00, AP-2219 New Year's Eve by 17:00, then AP-2207 spring break by Friday 12:00. The four low-risk moves can wait for one sign-off on Friday.
Why $95 and not $89?
At $89 the model expects 7.9 exit seats sold, 0.3 more than at $95, but $703 against $722: $19 less on this flight, about $870 across the 46. $89 matches Larkspur's exit-row seat. If you would rather match the market while the test runs, cap exit rows at $89.
Can a family ever pay to sit together?
No. The family seating promise is locked: each child 13 and under sits beside an accompanying adult at $0 on every fare. Only Claire Benton, with the director, can change a locked rule, and every change is logged.
How are the tests doing?
Two tests are running. T-38 Family bundle is ahead of control: attach 6.1% against 4.8%, +$1.42 per shopper, confidence 91%. T-39 is too early to call.
13 screens from the working solution, on its sample data. Pick one to see it large.
The limits your ancillary revenue management team sets for the agents, owned by the director, and the shopper-type inputs Counsel owns. Every change is logged.
| Setting | Default | Choose from |
|---|---|---|
| Director co-signs above | $25,000 estimated effect | Any dollar amount the director sets |
| Largest daily move | 20% per zone per flight, to the nearest dollar | Any percentage the director sets; a bigger move posts in steps |
| Checker gap that needs a person | $5 | Any dollar amount the director sets |
| Largest test cell | 20% of shoppers | 5%, 10% or 20% picked at approval, up to this limit |
| Longest test | 28 days | Any number of days the director sets |
| A feed is stale after | 6 hours | Event forecasts after 48 hours; a stale feed is not used |
| Remind before decide-by | 4 hours | Any number of hours |
| Inputs the agents may use | Trip context only | Locked; owned by Counsel |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the pilot · “estimated” = our estimate · “typical” = published figures (IdeaWorksCompany, November 2025: $157B global ancillary revenue in 2025, 15.7% of airline revenue against 9.1% in 2016). 3% of a $1.2B seat, bag and bundle line is about $36M a year, our arithmetic. Regulations named: 14 CFR 121.585, Part 382, Part 260. Corvane Airways, its rival Larkspur Air, and every person named on this page are fictional — characters and sample data in the working solution.
Airline Dynamic Ancillary Pricing is dynamic ancillary pricing for airline ancillary teams: airline dynamic pricing for paid seats, bags and bundles, set per shopper type and flight, inside rules people set. Eight agents read demand, propose a price for every paid seat zone, bundle and bag offer, check it against the guardrails, re-compute it and design a test cell. The ancillary pricing manager approves, caps or rejects, and the director co-signs above $25,000.
There is no family seating fee. Seat pricing optimization never touches the locked family seating rule: each child 13 and under sits beside an accompanying adult at $0 on every fare. The Guardrail keeper checks every price against it before a person sees it, the seat map can be viewed as a family sees it, and every test cell stops if any family is charged to sit together.
No. Prices are set per shopper type, built only from trip context: party size, children 13 and under, days to departure, trip length, fare family, channel, and loyalty tier for waivers only. Name, address, device, browsing history, past spend and any demographic are never used, and Counsel owns that list.
Every new price starts in a test cell — 5%, 10% or 20% of shopping sessions, assigned by session and never by person — with current prices as control and stop rules written first. The Test designer checks running tests every morning, and after departure the Outcome reviewer compares test with control and says what to keep.
Yes. Exit rows are never offered to a party with anyone under 15. 14 CFR 121.585 bars passengers under 15 from exit seats, and the airline's own rule keeps the rest of their party out of the exit rows too. Seat accommodations for passengers with disabilities stay at no charge (14 CFR Part 382). Both are locked rules the Guardrail keeper applies to every price.
The Fee refund watcher finds every passenger who loses a paid seat, chooses a new seat, keeps families together and prepares the refunds under 14 CFR Part 260. In the working solution a 737 MAX 9 to MAX 8 swap meant 14 refunds, $1,106, ready for one approval; nothing is refunded without a person.
No. The agents read, propose, check and explain. They never post a price, change a published bag fee, issue a refund or use personal data; a bag fee change goes to Distribution as a fee-change request. Nothing reaches shoppers without the pricing manager's yes and a reason code.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your fee tables, seat zones per aircraft, guardrails, approval limits and system connections — and it goes live once every quality gate has passed. First-time setup connects the systems, maps seat zones per aircraft and sets the first test.
One more point of revenue per flight, and no seat change posted without an analyst's yes.
1-min demoAirline Fare FilingEvery filing shopped before it publishes, so a mistake fare never reaches a shopper.
1-min demoCompetitive Fare MonitoringEach morning, the five rival moves worth money, with the captures that prove them.
1-min demoWe'll run Airline Dynamic Ancillary Pricing on a sample of your own flight-dates, seat maps and fee rules.