In partnership with
Hey, it's Jeffrey, back again!
My Amex Platinum renews in a couple of months, and the annual fee just jumped to $895, up from $695. That is a 29 percent increase, and just a lot to pay for a card in general.
I also carry a Chase Sapphire Reserve, so between the two cards I am already paying a lot in annual fees every year. I have dug into that redundancy before, including the issue where one of my three cards barely cleared its fee and the one where I worked out whether points were actually beating cash back. This time the question is narrower: with the Platinum's fee going up, is it still worth keeping at all?
Most years, I just let the renewal happen. The fee is charged; I shrug, pay it, and then move on. This year, I decided to take a closer look, especially since we have a Chase Sapphire Reserve card with a high annual fee.
One thing that made this easier: Amex lets you download a year-end benefit summary right from your account. It breaks down exactly which credits you used and how much you actually redeemed, so you’re not digging back through 12 months of statements by hand. I fed that summary, plus my full activity export, into Claude to build the comparison below.
Here's the prompt I used:
My credit card's annual fee is about to increase at renewal. I want to decide whether to keep, downgrade, or cancel before the new fee hits, based on real numbers, not the benefits page.
1. List every credit, perk, or benefit I actually use, with how often and how much. If my card issuer provides a year end benefit or rewards summary, use that as the source instead of manually combing through statements.
2. For any credit that requires spending I would not otherwise make, value it at zero, not face value.
3. Add a fair dollar value for non cash benefits I use in person, like lounge access, based on what I would pay out of pocket for the same access without the card.
4. Add up rewards earned on my real spend, valued at a realistic redemption rate, not the best case.
5. Subtract the new annual fee, not the old one, plus any authorized user fees.
6. Compare that net number to what the same real spend would earn on a flat 2 percent cash back card with no annual fee. Show me the comparison two ways: once counting only what shows up on my statement, and once including the non cash benefits.
7. Give me a verdict: keep, downgrade, or cancel, and tell me which single benefit is doing the most work to justify the fee.
Here is my situation:
Old annual fee: [$X]
New annual fee: [$X]
Authorized user fees: [$X]
Credits and benefits I use: [list them]
Approximate annual spend on this card: [$X]
Non cash benefits I use in person: [e.g. lounge access, hotel status]
How often I use each: [X times per year]
Here's what came up when I ran it on mine.
Here's what came up when I ran it on mine.
If I count every credit at face value, everything that ever hit my statement, I look great. Comfortably ahead of the new $895 fee, even with the authorized user fee added in.
But two of those credits, a retail credit and a hotel credit, were things I would not have bought or booked without the credit nudging me into it. That's not savings, that's spending with a coupon attached. Once I take those two out and count only the credits I'd have used anyway, the honest number flips. A flat 2 percent cash back card would have beaten my Platinum this year.
The card only wins again once I add something that never appears on any statement: the lounge access I actually use several times a year, for two of us. Counted at a fair value, that alone flips the comparison back in the Platinum's favor.
Amex Platinum | Flat 2% cash back card | |
|---|---|---|
Every credit at face value | about $440 (wins) | about $145 |
Excluding credits I wouldn't have used without them (Lululemon, Fine Hotels credit) | about −$370 | about $120 (wins) |
Adding the lounge access I actually use | about $230 (wins) | about $120 |
In plain terms: counted at face value, the Platinum wins by about $295 a year. Strip out the credits that were really just spending in disguise, and cash back wins instead, by about $490. Add the lounge back in, and the Platinum wins again, by about $110.
So the honest verdict for me is keep, but only because of one specific benefit. If I stopped using the lounge, or it stopped being available, I would drop this card the same week.
The best prompt engineers aren't typing. They're talking.
Power users figured this out early: speaking a prompt gives you 10x more context in half the time. You include the edge cases, the examples, the tone you want — because talking is fast enough that you don't skip them.
Wispr Flow captures everything you say and turns it into clean, structured text for any AI tool. Speak messy. Get polished input. Paste into ChatGPT, Claude, Cursor, or wherever you work.
89% of messages sent with zero edits. 4x faster than typing. Works system-wide on Mac, Windows, and iPhone.
Keep going? Want to go further? Try these follow-up prompts after you get your initial result:
Retroactive check: "Run the same test but with the higher fee applied retroactively to last year's spending. Would I have made the same keep decision a year ago?"
Downgrade path: "If I downgrade instead of cancel, what do I actually lose, and is the lower fee version still worth it for how I use the card?"
Build the habit: "Help me set a system so I re-run this test every year before my renewal date, not after the charge already hit."
AI tip worth trying this week: building a festival schedule with AI
I'm heading to Newport Folk Fest, a 3-day music festival, and I don’t know most of the bands playing. I gave an AI the full lineup and told it roughly what I usually listen to, then asked it to build me a day-by-day schedule of who to see and where, based on my taste rather than just who's headlining.
I don't have to stick to it, but it gave me a solid starting point instead of standing around during set times trying to decide on the fly. The same idea works for any multi-stage event, conference, or festival with more happening at once than you can actually plan for yourself.
One quick note: This newsletter is for educational purposes only and does not constitute financial advice. I'm not a financial advisor, just someone sharing ideas and tools I've found useful. Use what works for you, skip what doesn't, and always do your own research. Some links may be affiliate links or sponsored content for which I may receive compensation.