How to Make Your Tours Bookable by AI Agents: An MCP Primer for Travel Suppliers
AI agents are learning to book. Here is what a tour supplier should put on the menu for them, what to leave off, and how to start without betting the season.
Field Notes · The Nova Adventure blog
Notes from the trail between travel and software
Practical, opinionated writing for people who run trips: how AI agents book, how a day gets routed, what predicts a cancellation, and where the hours in booking operations actually go.
All articles
The Itinerary Is a Constraint Problem
Every itinerary that falls apart on day two was already infeasible on day zero. Here is how to see the math hiding inside an ordinary trip plan.
Pacing: The Hidden Variable in Itinerary Personalization
Travelers rarely complain about the wrong museum. They complain about the day that was one stop too long.
Getting Cited by AI Assistants: A Visibility Playbook for Independent Travel Brands
Assistants recommend what they can understand, verify and quote. Most independent travel brands give them remarkably little to work with.
Predicting Cancellations Before They Happen
A cancellation score is only worth building if someone does something different on Tuesday because of it.
Draft, Don't Decide: Keeping Travel Advisors in the Loop
The point of AI in an advisory business isn't fewer decisions by people. It's better ones, made sooner, with the busywork already done.
Disruption Playbooks: Re-planning a Trip Before the Shuttle Leaves
Most disruptions announce themselves hours in advance. The operators who handle them well have already decided what they'll do.
First-Party Preference Profiles Travelers Actually Trust
A preference profile is a promise about what you will do with what a traveler tells you. Here is a design that keeps the promise small, visible and useful.
Shaping Shoulder-Season Demand With Predictive Signals
Discounting is what you do when you have run out of time. Predictive signals buy that time back. Here is how to spend it, week by week.
Accessible Itineraries by Design
Accessibility isn't a property of a venue. It is a property of a particular person's day, and that day can be designed.
Low-Carbon Routing Without the Guilt Trip
Travelers don't buy penance. They buy better trips, and on the ground the lower-carbon trip is usually the better one.
Where the Hours Go: An Anatomy of Booking Operations
Ask a booking desk where the week goes and you will hear 'email.' Log it in ten-minute blocks and the answer gets more specific, and much more fixable.
Group Trips Are Negotiations
A group doesn't have a preference. It has several, plus a memory of who gave in last time. Plan for both.
How We Evaluate an Itinerary Generator Before Travelers See It
A generated itinerary can be fluent, on-brand and impossible. Here is the protocol we run before any of them reach a traveler.
Supplier Truth: Content Mapping, Deduplication and the Data Behind Personalization
Before a recommender can know what a traveler wants, it has to know what actually exists. Here is how to turn five messy feeds into one true record per real-world thing.
Twelve Questions DMOs Should Ask Before Buying an AI Trip Planner
Every demo looks good. These questions separate a planner that serves your destination from one that only borrows its name.
Phantom Trails and Closed Cafés: Grounding Travel AI in Live Data
A recommendation is a claim about the world at a particular time. Treat it that way, and most phantom venues never reach a traveler.
When an AI Concierge Should Stay Quiet
An in-trip assistant is judged as much by the messages it doesn't send as by the ones it does.
High-Desert Lessons: What Northern New Mexico Teaches About Context
Out here, a plan that looks right on a map can be wrong by early afternoon. What the high desert has taught us about modeling context.
Agent-Safe Payments: Letting AI Agents Book Without Losing Control
An AI agent holding a card number is a liability. An agent holding a scoped mandate, a locked quote and a human checkpoint is a sales channel.
Want these ideas applied to your operation?
A 45-minute discovery call is the fastest way to find out which of these notes matter for your business, and which can wait.