I'm a points expert. AI still found money I left on the table in Rome.

Reading Time: 5 min

I consider myself an expert at booking award travel and maximizing my earnings on purchases. However, during my four-night trip to Rome in February, I was curious if I left any opportunities on the table.

I told ChatGPT how much I spent on airfare, hotels, excursions, dining, and other expenses during my trip, along with which cards I used for each purchase, and asked it to audit my trip. This is what it said.

To get feedback on how “well” I did with my trip, I fed the following into ChatGPT:

Travel to Rome:

• We took a train from Milan to Rome, and I booked it using my Capital One Venture X card. It cost $128.89.

Travel leaving Rome:

• I booked my flight from FCO to MEX via CDG using 30,000 Air France-KLM Flying Blue miles, plus €229.41 in taxes and fees paid with my Platinum Card (premium economy). I then paid $388 using my Platinum card to upgrade to business class.
• I booked my partner's flight from FCO to EZE using 55,500 Air France-KLM Flying Blue miles, plus €186.77 in taxes and fees paid with my Platinum Card (economy).

Hotel:

• I booked The Hoxton Rome through the Amex Travel Portal. It cost $798.29, and I received $300 back thanks to the Platinum Card's hotel credit, awarded semi-annually.
• I paid the tourist tax of $87.34 using my Platinum Card because I thought it was charged by the hotel, but it appears on my statement as "other," so it only earned 1 point per dollar.

Other expenses:

• I charged $121.05 in dining to my Citi Strata Elite.
• I paid $106.84 for a tour of the Colosseum with my Capital One Venture X card.
• I paid $129.03 for a new carry-on suitcase to avoid an overweight baggage charge on Air France using my Capital One Venture X.
• I bought a new leather belt for $47.35 using my Capital One Venture X.
• I spent $358.06 on Uber and taxis with my Capital One Venture X card.

Your points strategy is optimized. Is your wallet?

High-value travel requires high-performance budget optimization. ​MeetFranklin.ai is the conversational AI that turns your spending data into a strategic advantage.​ Instead of wrestling with spreadsheets, just ask: “How much did I spend on dining in Tokyo?” or “Can I afford that business class upgrade today?” Franklin automatically identifies forgotten subscriptions, visualizes your cash flow with beautiful Sankey diagrams, and alerts you to savings opportunities — all with bank-level security. It’s the smarter way to ​manage the "cash" side of your travel hacking​. Points Path Insider readers can start their ​30-day free trial​ today — no credit card required.

Get started for free

Here’s how I performed in my audit:

It told me that I got solid value from my flight redemptions, and that using my Platinum Card for taxes and fees was the right choice, thanks to the 5x Membership Rewards points on airfare. However, it flagged that paying to upgrade was a red flag because I didn’t earn any additional miles or elite credits for the upgraded cabin.

The hotel audit mentioned that by booking through Amex Travel, I lost the opportunity to earn Accor ALL points or work toward status. However, since it is a member of The Hotel Collection, that tradeoff was likely worth it.

One of the biggest mistakes I made was in paying the tourist tax, which would’ve earned more miles if I had used the Capital One Venture X instead of the Platinum Card. It recommended asking the hotel to bundle it into my final folio next time if I want to pay with the Platinum Card.

It also flagged that I didn’t link a rail loyalty program to my train booking to Rome, and that I could’ve booked through a shopping portal to stack my rewards. I should’ve also used a portal to earn more on my Colosseum tour. However, I used the right card for my dining purchases.

For ground transportation, ChatGPT recommended using a card that earns 3x points on rideshare purchases, such as the American Express Green Card or the Wells Fargo Autograph Card. Unfortunately, I don’t have one of those cards, but I do have the Apple Card, which would’ve earned 3% cash back as long as I used Apple Pay.

In its estimation, I had around $230–500 in missed upside. I think I did well overall, but it's a reminder to always use shopping portals to double-dip on rewards. For longer, more complex trips, this can be a fun experiment to see how to optimize future travel.

In case you missed it...

9 ways to instantly upgrade your hotel room

Want to run this analysis on your own trip? Copy and paste the prompt below into ChatGPT (or your AI of choice), and fill in your details:

I want you to audit a recent trip for missed points and miles opportunities. For each purchase below, evaluate whether I used the optimal card, whether I should have stacked a shopping portal or loyalty program, and whether there was a better redemption or earning strategy I missed. At the end, give me a rough estimate of the total value I left on the table

FLIGHTS

• Origin → destination: [ ]
• Cabin: [ ]
• How I booked it (cash, points + which program, or portal): [ ]
• Miles used + co-pay amount: [ ]
• Which card I paid with or used for taxes/fees/cash portion: [ ]

HOTEL

• Property name: [ ]
• Total cost: [ ]
• How I booked (direct, OTA, portal, or transferred points): [ ]
• Which card I used: [ ]
• Any credits or perks I applied: [ ]

GROUND TRANSPORTATION

• Trains, rideshares, taxis, rental cars: [ ]
• Total spent for each: [ ]
• Which card I used for each: [ ]

ACTIVITIES AND TOURS

• What I booked: [ ]
• Total cost: [ ]
• Which card I used: [ ]
• Whether I went through a portal: [ ]

DINING

• Total spent: [ ]
• Which card(s) I used: [ ]

OTHER PURCHASES

• Anything else charged during the trip — shopping, baggage fees, tourist taxes: [ ]
• Total cost: [ ]
• Which card(s) I used: [ ]

CARDS IN MY WALLET

• [List every rewards card you carry, even ones you didn't use on this trip, so the AI can flag better options you already had access to.]

Flag any purchase where I used a suboptimal card, missed a portal, skipped a loyalty program signup, or made a redemption choice that cost me earning potential. Be specific about which card or program I should have used instead.

Samsonite's lightweight hardside holds just about anything

At just 6.7 lbs empty, the Samsonite Winfield 2 Hardside touts a lightweight polycarbonate shell with a scratch-hiding brushed pattern, 360-degree spinner wheels, TSA-approved lock, and an organized interior.

Get it at Amazon