Free AI Workshops
AIHA Academy — 8 free live AI workshops for hospitalityFrom AI basics to building your own agents and appsLive online · One hour each · Recordings included
Register free
Back to Insights

2026-10-07 | Rafat Ali

Convenience vs. Productivity Is Travel AI's Real Test

Skift's Rafat Ali argues travel AI should be judged by how much responsibility travelers can hand over, not how many steps an agent automates, and that the real value sits in staying responsible after the booking.

This insight summarises Convenience vs. Productivity Is Travel AI's Real Test, a Connecting the Dots column by Skift founder Rafat Ali, published on Skift on 7 October 2026.

The wrong yardstick. Ali borrows a distinction from Ben Thompson's "The iPhone's Last Stand": enterprises pay for employee time, so productivity sells at work, but consumers mostly want tasks and coordination burdens to move somewhere else. "An agent that has to be watched, corrected, and checked after every step has not removed much work." For travel AI, the measure is how much responsibility the traveler hands over, not how many steps an agent automates.

A four-step ladder of responsibility:

  1. Answer: explain the options.
  2. Compare: narrow the choices.
  3. Execute: book, with the traveler's approval.
  4. Assure: stay responsible after the transaction by monitoring flights, protecting loyalty benefits, chasing refunds and repairing disruptions.

Ali places the real value in the fourth step, because "a confirmation number does not mean the problem is finished." When a flight is cancelled, a useful agent works out what the rebooking does to the family's seating, the hotel stay, the airport pickup and any refund, so the traveler does not have to become their own disruption manager.

Convenience is selective. Expedia CEO Ariane Gorin said at Skift Global Forum that she does not expect a large share of bookings to become one continuous conversation: planning is multi-turn, often involves other people and sometimes benefits from slowing down. Expedia moved from one universal agent to narrower agents for specific parts of planning and shopping.

Where to hand over control. Routine work such as re-entering passenger details or filing a refund can be automated. Splitting a family across flights or giving up a loyalty benefit should still need approval. Reliability, privacy, access to underlying services and human fallback remain the constraints.

Why this matters for hotels. Guests will judge AI by the work it removes after the booking as much as before it. Hotels that can confirm, modify and repair stays reliably through agents and partners will be the easiest to trust with the "assure" step.

Read the full article on Skift →

Stay Updated

Enjoyed this insight?

Get more AI hospitality research and trends.

For Contributors

Submit article for consideration

Share an article, perspective, or industry analysis for possible inclusion in the Alliance Insights section.