Stay up
to date
New articles. New card offers. Clean strategy delivered straight. No noise, just the signal.
How to decide whether an annual fee is worth paying
A four-line calculation that tells you whether to keep, downgrade or cancel a card - and why 'it pays for itself' is usually the card talking.
.jpeg)
Losha Nazarov
Losha Nazarov I enjoy finding smart ways to earn more rewards from money I'm already spending.
A premium card’s marketing works by addition. It lists a $300 travel credit, a $200 hotel credit, $200 of rideshare credit, lounge access “worth” $469, and a $100 fee credit, arrives at $1,269 of value, sets that against a $695 fee, and concludes that you are up $574 before earning a single point.
Every individual number there may be technically accurate. The total is still fiction, for three reasons that the addition carefully hides.
The three things the marketing total ignores
It counts credits you will not fully use
A $200 annual credit paid as $200 is worth $200. The same $200 paid as $16.67 a month, on one specific rideshare app, expiring monthly, is worth considerably less — because some months you will not use it, and the unused portion does not roll over. That is not an accident of design. It is the design.
Our card finder handles this with a “realistic” setting that discounts each credit by the share a typical cardholder actually claims. Monthly-expiring credits on narrow merchant lists get discounted hard. Annual credits that apply automatically to any travel purchase barely get discounted at all. You can also set it to “use all” and see the marketing version, which is instructive mostly as a measure of how much of the pitch depends on perfect behaviour.
It counts benefits you would never have bought
Lounge access is quoted at the price of an annual membership nobody pays. If you fly four times a year, you are getting four lounge visits, and a lounge visit is worth roughly what you would otherwise have spent on an airport sandwich and a drink — call it $35. Four visits is $140, not $469.
The rule: value a benefit at what you would have paid for the thing it replaces, never at its sticker price. A hotel status you will use for two nights a year is worth two free breakfasts, not the status’s notional value.
It counts credits that require spending to unlock
A “$100 credit on a $500 hotel booking” is a 20% discount on a $500 purchase, not $100 of free money. If you were not going to book a $500 hotel, it is worth zero. If you were, it is worth $100. Treat these as discounts and ask whether the underlying purchase was happening anyway.
The four-line calculation
Once a year, on the month your fee posts, write down four numbers.
| Line | What to write down |
|---|---|
| A | Rewards this card earned me last year, valued at how I actually redeem |
| B | Rewards a free 2% card would have earned on the same spending |
| C | Credits and benefits I actually used, valued at what I would have paid otherwise |
| D | The annual fee |
Then: (A − B) + C − D.
If the result is positive, keep the card. If it is negative, downgrade it. That is the entire decision.
Line B is the one people leave out, and leaving it out is what makes bad cards look fine. The question is never “did this card earn me more than its fee?” — almost any card clears that bar on enough spending. The question is “did this card earn me more than the free card I would otherwise have used?” Only the difference is attributable to the fee.
Worked example. A $395 card earned 48,000 points on $32,000 of spending. At 1.6¢ that is $768 (A). A 2% card would have earned $640 (B). You used $300 of travel credit and got $130 of anniversary miles (C). So: (768 − 640) + 430 − 395 = +$163. Keep it — but notice that the rewards advantage was only $128. The credits are doing almost all the work, which means the card is one benefit cut away from being a loss.
Downgrade, do not cancel
When the number comes out negative, cancelling is rarely the right move. Closing an account eventually removes it from your average account age, which is about 15% of your credit score, and it can raise your overall utilization by removing the credit line.
Most issuers will “product change” a premium card into a no-fee card in the same family. Same account, same age, same credit line, no fee. It is a short phone call and the agent will usually offer it before you have finished explaining.
Two caveats. Ask about the timing — most issuers refund the fee if you change within about 30 days of it posting. And ask whether points transfer; on some products, downgrading strands transferable points in a cash-back-only account where they lose most of their value, so move or spend them first.
The retention offer
Before you downgrade, call and say plainly that the fee is hard to justify and you are considering closing the account. Issuers frequently respond with a retention offer: a statement credit, bonus points, or a fee waiver.
This is not a trick and you are not being rude. Retention offers exist as a budgeted line item because acquiring a new customer costs more than keeping one. Be honest, be polite, and be willing to actually downgrade if there is no offer — the leverage is only real if you mean it.
Fees that are almost always worth it, and almost never
Usually worth it: a $95 fee on a category you spend heavily in and where the multiplier gap over a free card exceeds roughly $150 a year. An airline card whose free checked bag covers a family that checks bags twice a year. A hotel card whose annual free-night certificate reliably exceeds the fee.
Usually not worth it: any premium card bought for a signup bonus you will not repeat. A lounge card if you fly fewer than about eight segments a year. Any card whose value case rests mainly on monthly-expiring credits — unless you have genuinely proved to yourself, over a full year, that you claim them.
The honest version of the premium card pitch is this: it is a subscription to a set of benefits. Subscriptions are worth it when you use them. Nobody would argue about a gym membership by adding up the retail price of every machine.
Open the card finder to run lines A through D on your own numbers — every result shows net value after the fee, with credits discounted to what a real person claims.