How AI helped me save $1,000 on a family award trip

Reading Time: 4 min

I recently used AI to help solve one of my more complicated award bookings. Where I normally would have used scratch paper and an Excel sheet, I fed my constraints into ChatGPT and asked it to recommend the lowest out-of-pocket booking strategy.

I was booking a Thanksgiving trip on Alaska Airlines for my family of four: two adults and two kids. I needed to find the best order to use miles, companion fares, transfers, and cash.

I had two Alaska companion fares that were approaching expiration. Companion fares can only be used on paid tickets, not award tickets. They allow a second traveler to fly for about $121 plus taxes and fees, and they can be used on either a round-trip or a one-way booking.

That meant I had to decide where the companion fares would create the most value or if I should use them at all. There can be instances where the fare per person on Alaska Airlines is lower than $121, in which case you should not use the companion fare.

I also had Alaska points spread across four different Atmos Rewards accounts. Some were in my account, some were in my spouse’s account, and some were in accounts belonging to relatives.

There were enough points to be useful, but not enough in any one account to book the trip cleanly. Alaska allows members to transfer Atmos rewards points between accounts, but charges roughly 1 cent per point plus a $25 processing fee. That cost had to be built into the analysis.

Then there were the family logistics. We were traveling with two young children, and I wanted each child booked on the same reservation as a parent. Alaska does allow minors to travel on separate reservations that are linked to an accompanying adult's reservation, but I preferred to keep each parent-child pair on the same booking to avoid any complications. So I couldn’t optimize four individual tickets. They had to be in pairs.

The outbound and return flights were priced very differently. The outbound flights were a much better points redemption than the return flights. The return flights were more expensive in cash, which made them better candidates for the companion fares.

So I had several variables interacting at once: companion fares, mileage balances, transfer fees, expiration dates, cash prices, award prices, and parent-child reservation rules.

In case you missed it...

7 Father's Day gifts a frequent-flyer dad will use

I put all of those details into ChatGPT and asked it to find the lowest out-of-pocket way to book the trip. I didn’t even have to write much — I just dumped in screenshots of flight costs and point redemptions, terms of the companion pass, terms of mileage sharing, and points balances. I also briefly explained the parent-child situation.

ChatGPT compared round-trip bookings against one-way bookings, calculated transfer costs, evaluated different ways to use the companion fares, and accounted for the parent-child reservation requirement. It also factored in the reality that some miles would need to be consolidated before any booking could happen.

Here’s the prompt I fed into ChatGPT:

Factoring in the details below, find the booking strategy that minimizes my total out-of-pocket cost and tell me exactly what steps to take to make the bookings.

Here are my flight options: [ ]
Points balances: [ ]
Companion fares: [ ]
Terms and conditions for transferring points between flyers:
Booking constraints: [ ]

ChatGPT’s recommended strategy was to redeem points for all four outbound flights, where the redemption value was strongest, and save the companion fares for the more expensive return flights.

I had to do the booking carefully. The trip was booked as two separate one-way itineraries, with award tickets in one direction and companion fares in the other. After accounting for transfer fees and taxes, the strategy reduced my out-of-pocket cost by roughly $1,000 compared with simply paying cash for all four tickets.

Using a companion fare changed where I wanted to use rewards, and using rewards affected whether paying a transfer fee made sense. Booking a child with one parent affected which account needed the miles.

In short: Lots of decisions were conditional, and I needed to consider every combination. I mostly wanted confidence that I wasn’t overlooking a better combination.

A similar prompt can help you with more travel and rewards decisions than you realize! Any time you're juggling companion fares, expiring credits, family mileage balances, transfer fees, multiple loyalty programs, weird terms and conditions, expiration dates, or cash-versus-points decisions, AI can help evaluate the tradeoffs and recommend the lowest-cost path. The more moving pieces involved, the more time it can save.

This travel essential silences engine roar and busy terminals on a budget

Anker Soundcore Q30 wireless headphones create a personal sanctuary with hybrid noise cancelling, specialized transport modes, and ultra-soft protein leather earcups. Their sound rivals expensive brands and they stay comfortable for hours.

Get it at Amazon