Skip to content

Should the Sleeper Bus Price Like an Airline?

The Prompt

Your client runs 120 premium sleeper buses on intercity routes like Mumbai–Goa and Bengaluru–Hyderabad, selling at fixed prices per route (~₹1,400 average). Occupancy swings from 55% midweek to 100% sold-out-with-waitlist on Fridays, Sundays, and holidays. The CEO asks: should we adopt airline-style dynamic pricing, and what would it earn us?

Structure & Hypothesis

Two leaks, opposite fixes — and the math tier under each. The honest part is subtracting the ₹6k dilution.

Analysis & Data

interviewer

Size the prize. A bus has 36 berths; assume 40 peak departures and 80 off-peak departures a week across the fleet… actually, let's keep it per-bus-per-week: 3 peak and 4 off-peak departures.

candidate

Peak: 3 departures × 36 berths, currently all at ₹1,400. With a stepped curve — first third at ₹1,400, middle at ₹1,800, last third at ₹2,300 — average ≈ ₹1,830, +₹430/berth: 3 × 36 × 430 ≈ +₹46,400/bus/week, and the waitlist says volume holds. Off-peak: 4 × 36 at 55% = 79 filled; an advance-purchase ₹999 fare could lift occupancy to ~70% — 22 extra berths × ₹999 ≈ +₹22,000, minus dilution from full-fare bookers who'd have paid ₹1,400 anyway, say 15 × ₹400 = ₹6,000 — net ≈ +₹16,000. Total ≈ ₹62,000/bus/week, ~₹39 crore annualized across 120 buses — roughly a 20% revenue lift at near-zero marginal cost.

Counts the dilution honestly. Most candidates forget that off-peak discounts also reach people who would have paid full fare.

interviewer

And the fairness backlash?

candidate

Three design rules: cap the peak-to-floor ratio at ~2.3× (airlines run 5–10× and get hated for it); frame everything as early-bird savings off a higher anchor rather than surge on a base; and keep prices fixed once booked — no repricing, no airline-style fare classes visible. Buses sell to families, not expense accounts; perception discipline is worth a few rupees of theoretical yield.

Recommendation

Recommend to the CEO

  • Adopt stepped, occupancy-triggered pricing: 3 fare steps per departure, peak ceiling ~2.3× the off-peak floor — not continuous airline-style surge.
  • Frame as early-bird discounts from a raised anchor; never let a customer watch a price rise mid-booking.
  • Pilot on two routes for eight weeks; measure revenue per departure, direct-app share, and review sentiment before fleet rollout.
  • Use the off-peak ₹999 fare to attack rail-substitutable midweek demand, and push it through the direct app only — protecting aggregator rank at peak.

Key Takeaway

What this case teaches

A waitlist is a price signal, not an ops problem. Dynamic-pricing cases are 30% arithmetic and 70% mechanism design — caps, framing, and channel handling decide whether the math survives contact with real customers.