Fare families were designed to make airline pricing legible to a customer staring at a long search-results list, and to give revenue managers a small set of price points to defend in meetings. Both problems still exist. But the form-factor — three or four named bundles, refreshed quarterly — is the wrong shape for travel retailing in 2026.
Travelers no longer compare on price alone. They compare on what's included, what's optional, and how the bundle maps to the trip they're actually taking. Static bundles can't keep up. The teams winning right now are running thousands of micro-bundle experiments per month — not because they enjoy the operational overhead, but because the lift compounds.
The replacement is not 'no bundles.' It's goal-driven bundles: a system that knows the target metric (attach, RPK, NPS), proposes candidate bundle structures, and tests them with margin and policy guardrails. The bundles still feel like bundles to the traveler. The difference is that there are forty live versions, not four.
What this looks like in practice: airlines we work with run between fifty and four hundred concurrent bundle experiments. Each experiment is a small variant — a different inclusion, a different price point, a different visual treatment. Most don't move the needle. The ones that do compound week over week, and within a quarter the average ancillary RPK is 18-24% higher than the pre-experiment baseline.
What this requires: a merchandising system that treats bundles as data, not as code; a pricing system that can value a bundle at composition time, not at list price; and an attribution system that can tell you which experiment moved which dollar.
We expect the named-fare-family construct to survive at the surface level — travelers like the cognitive shortcut. What's changing is everything underneath: the bundles being shipped under those names will be dynamic, and increasingly per-segment, per-route, and per-traveler.