(505) 303-9060Book a discovery call

Solutions · Experiences & activities

Fewer empty seats, fewer 4 AM phone trees

Balloon flights, river trips, food walks, cooking classes: the product lasts a few hours, and everything around it — weather calls, rebooking, a dozen sales channels — never stops. Nova takes on the parts that don't need your guides.

Book a discovery call Make your tours agent-bookable
A kayak rack where faint ghost kayaks are being filled by paper boats drifting down the river.
Field noteEvery empty seat is perishable. The goal is the right booking reaching the right seat in time.

4:30 AM, at the launch field

The pilot is reading wind data in the truck. The forecast said marginal, and marginal is exactly what it is: calm at the surface, a stronger layer a few thousand feet up, moving the wrong way. Twenty-two passengers are booked for the 6:40 launch. They came through your own site, two resellers and a hotel concierge desk, and three of them fly home tomorrow afternoon.

At 4:45 it’s a no-go. That decision takes fifteen minutes, and it’s the right one.

What follows takes until lunch. Someone in the office starts texting from a spreadsheet. The reseller bookings have no direct phone number. One couple found you through an AI assistant reading a third-party listing that gave the wrong meeting point, so they’re already standing in a dark parking lot on the other side of town.

By 9 AM, fifteen people are rebooked, four want refunds and three haven’t answered — including the pair flying out tomorrow. Meanwhile, Tuesday’s departure is half full again, and nobody has had time to ask why.

Side by side

What changes with Nova

The pilot still reads the wind. Everything before and after that decision gets lighter.

TodayWith Nova
The go / no-go callMade at 4:45 AM, then followed by an hour of texts and calls.Still made by your pilot or lead guide. One confirmation triggers the whole rebooking playbook.
RebookingTwenty-two passengers, four channels, one spreadsheet.Each guest gets rebooking options in their own channel and language. Only the awkward cases — gift vouchers, reseller bookings — reach the office.
Soft daysTuesday departures run half-empty, or get cancelled at the last minute.Signals forecasts weak departures early, so you can consolidate, promote or reschedule with real notice.
DistributionResellers and listing sites own the customer — and take the commission.Agent Gateway lets AI assistants find and book you directly, under your own policies and prices.
Pre-trip questionsWhat to wear, where to park, can my nine-year-old come — answered by hand, one at a time.Concierge answers from your own facts and hands off anything it shouldn't guess.

Where we'd start

Weather-dependent operators usually start with disruption handling. Everyone else tends to start with distribution.

How it works

A sample workflow: a scrubbed sunrise launch

  1. 01

    The evening before

    T–10 hours

    Forecast data for the launch window arrives automatically. A marginal forecast is flagged to the pilot with the passenger list and tomorrow's open slots.

  2. 02

    The call

    T–2 hours

    The pilot checks conditions and makes the go / no-go decision. Nothing about that changes.

  3. 03

    One confirmation

    A no-go triggers the playbook agreed in advance: which dates to offer, what to say, which guests get a phone call instead of a message.

  4. 04

    Guests choose

    Concierge messages each party with three rebooking options, plus a refund link where your policy allows. Most choose before breakfast.

  5. 05

    The office takes the odd ones

    Reseller bookings, gift vouchers and the family leaving town tomorrow land in a short queue with the context attached.

  6. 06

    Signals learns

    Scrub rates by month and site feed next season's forecasts — and the question of whether Tuesday departures are worth running.

The pilot still makes the call. The difference is that at ten to five in the morning, the call is the only thing he has to do.

Owner, hot-air balloon and sunrise-tour operator in the high desert

Related work: knowing early who won't show

Our closest case study comes from hospitality rather than activities, but the problem underneath is the same: knowing early which bookings are wobbling, and doing something useful about it. Read how a collection of desert lodges used cancellation prediction and shoulder-season packages — the same signals apply to departures and launches.

Read the case study

Questions

Frequently asked questions

Does the AI decide whether a tour runs?

No. Weather and safety calls stay with your pilots and guides. Nova prepares the information before the decision and runs the playbook after it.

Can AI assistants book us directly, without a reseller?

Yes, through Agent Gateway: an MCP server and agent-safe checkout that expose your products, availability and policies to AI assistants on your terms. You keep the customer relationship and the margin. Our primer on agent-bookable tours explains how it works.

We sell through several resellers. Will this cause double bookings?

Availability is read from your reservation system as the single source of truth, so every channel — human, reseller or agent — draws from the same inventory.

Is this only for weather-dependent activities?

No. Food walks, gallery tours and cooking classes use Route Engine for timing and opening windows, and Signals for demand on soft days. Weather is just where the pain is loudest.

What do we need to get started?

Access to your reservation system, a season or two of booking history, and an afternoon with whoever handles rebooking today. A Discovery Sprint maps the rest.

Related

Keep exploring

Show us your last scrubbed morning

Walk us through the messages, the calls and the spreadsheet. We'll show you which parts Nova would handle and which stay with your team.

Book a discovery call Read the agent-ready checklist