American Express Gold at $325: Where Dining and Grocery Multipliers Pay Off

The Gold is designed around a specific cardholder profile: households whose discretionary spend skews to dining and groceries. The two highest-volume monthly spend categories for many US households are food at home (groceries) and food away from home (restaurants and delivery). A card that earns 4x on both, capping only the supermarket category, captures more of the household spend at a bonus rate than most competitor cards.

This positioning differs from the Sapphire Preferred (travel-and-dining oriented) and the Sapphire Reserve (travel-portal oriented). The Gold is more correctly described as a food-and-flights card. The 3x airfare-direct earn covers flights; the 4x dining and 4x supermarket cover food. Other spend categories earn 1x baseline, which is uncompetitive against flat-rate 2 percent alternatives.

The implication: cardholders whose food spend dominates the wallet captures most of the Gold's value. Cardholders whose spend is more balanced across non-food categories leave value on the table; the 1x baseline drags down the overall earn rate. The Gold is a specialised tool. The break-even maths reflects this.

Per the product page, the Gold's published earn structure:

• 4x on restaurants worldwide (including takeout and delivery in the US)
• 4x on U.S. supermarkets, on up to $25,000 of supermarket spend per calendar year; 1x thereafter
• 3x on flights booked directly with airlines or with Amex Travel
• 2x on prepaid hotels booked through Amex Travel
• 1x on everything else

The supermarket cap is the binding structural constraint. At $25,000 of supermarket spend per year (a recent reduction from the older $50,000 figure for some product variants), the 4x rate produces 100,000 points; anything beyond $25,000 in the calendar year earns at 1x. The supermarket categorisation is strict: Walmart Supercenter usually does not code as supermarket; warehouse clubs typically do not; specialty food stores typically do.

The restaurant earn is uncapped, which is the Gold's most aggressive feature for high-dining-spend cardholders. A cardholder spending $15,000 per year on restaurants and delivery earns 60,000 MR points on that category alone, valued at $900 at a 1.5 cpp transfer redemption.

Worked example: a cardholder with $40,000 annual spend, $10,000 dining, $8,000 supermarket, $5,000 direct-airline flights, and $17,000 baseline, earns: (10,000 + 8,000) × 4 + 5,000 × 3 + 17,000 = 72,000 + 15,000 + 17,000 = 104,000 points. At 1.5 cpp via transfer, that is $1,560 of annual reward value.

The Gold's credit stack, per the 2024 revision:

• $120 annual Uber Cash credit, issued as $10 monthly. Use-it-or-lose-it monthly.
• $120 annual dining credit, issued as $10 monthly statement credit on select dining partners (Grubhub, Five Guys, Goldbelly, and others). Use-it-or-lose-it monthly.
• $84 annual Dunkin' credit, issued as $7 monthly. Use-it-or-lose-it monthly.
• $100 Resy credit, issued as $50 in two periods (January-June and July-December). Use-it-or-lose-it.

Nominal total credit value: $424. Realistic value depends heavily on the cardholder's natural usage of these specific merchants. We compute three honest scenarios:

The unaligned cardholder who tries to engineer usage often spends additional dollars to claim the credit. A cardholder who buys a $10 Dunkin breakfast monthly to claim the $7 credit nets $3 of value but has spent $10 on coffee they may not have wanted. The honest valuation in this case may be zero or negative, not $7.

The aligned cardholder who already buys at these merchants captures real value. The Uber Cash redeems against rides the cardholder would have taken anyway. The dining-partner credit covers a Grubhub order the cardholder was going to place. The Dunkin credit covers the morning coffee the cardholder buys regardless. In this case the credit is worth nominal value.

The aligned cardholder pays a negative effective fee; Amex is paying them for holding the card before any rewards are considered. The partial cardholder pays $75 effective, which is comparable to Sapphire Preferred at the no-credit baseline. The unaligned cardholder pays $225, which requires substantial reward earn to clear.

For the partial-usage cardholder ($75 effective fee) at 1.5 cpp:

For the unaligned cardholder ($225 effective fee):

The unaligned cardholder needs roughly $3,750 of 4x-category spend to clear break-even before earning any net value. Most cardholders clear this in dining and supermarket spend alone. The card is unlikely to net-negative for active cardholders even with poor credit alignment, but the margin is much smaller and the case for holding the card weaker.

At 2.0 cpp (MR transfer to airline sweet-spot redemptions), the break-even points fall by 25 percent. The aggressive cpp matters more for unaligned cardholders, where the fee recovery is the binding constraint.

For most cardholders the supermarket cap is irrelevant; median grocery spend lands well below $2,000 per month. For high-volume cardholders the cap is binding. A family of four with $2,500 monthly grocery spend ($30,000 annual) earns 4x on the first $25,000 (100,000 points) and 1x on the remaining $5,000 (5,000 points), totaling 105,000 supermarket points instead of the 120,000 they would earn if the cap did not exist.

The lost-earn at the cap is 3x on the over-cap spend. At 1.5 cpp, the cost of the cap is 3 cpp times $5,000 of over-cap spend equals $150 of foregone value annually. For a $40,000 grocery cardholder, the cost is 3 cpp times $15,000 equals $450 of foregone value.

Cardholders hitting the cap have alternatives. The Bank of America Customized Cash Rewards card earns 3 percent on grocery store and wholesale club purchases on up to $2,500 per quarter ($10,000 annually) at no fee, which captures a different segment of the over-cap spend at a different rate. The Citi Custom Cash earns 5 percent on the cardholder's top category each statement, capped at $500 per cycle, which may or may not be grocery in a given month. The Amex Blue Cash Preferred earns 6 percent on US supermarkets up to $6,000 per year, then 1 percent thereafter, at $95 annual fee.

For households spending well above the Gold cap, holding the Blue Cash Preferred as a secondary supermarket card may capture more total value on grocery spend than relying on Gold alone. The two-card portfolio handles the segmentation that one card cannot.

The Amex Blue Cash Preferred ($95 fee) earns 6 percent on US supermarkets up to $6,000 per year, 3 percent on US gas and select streaming, 1 percent elsewhere. On the dining axis, Blue Cash Preferred earns 1 percent (no dining bonus).

For a cardholder with $6,000 supermarket spend and modest dining ($3,000 per year):

• Blue Cash Preferred: $6,000 supermarket × 6% = $360 + $3,000 dining × 1% = $30, total $390 cashback. Minus $95 fee = $295 net.
• Amex Gold: $6,000 supermarket × 4x = 24,000 + $3,000 dining × 4x = 12,000, total 36,000 MR. At 1.5 cpp = $540 reward. Minus $75 effective fee (partial usage) = $465 net.

Gold wins by $170 at this profile. But this assumes the partial-usage credit scenario; if the cardholder is in the unaligned scenario ($225 effective fee), Gold net falls to $315, narrower lead. At Unaligned plus 1.25 cpp portal redemption, Gold net falls to $215, losing to Blue Cash Preferred.

For dining-heavy cardholders, Gold wins more decisively. For supermarket-heavy cardholders, the comparison narrows. For both cardholders, the credit-usage alignment is the swing variable.

The Platinum's $695 fee dwarfs the Gold's $325. The Platinum's earn structure prioritises 5x on airfare (booked direct or through Amex Travel) and 5x on prepaid hotels through Amex Travel, with 1x baseline elsewhere. For a dining-and-grocery cardholder, the Platinum offers no bonus on the cardholder's primary spend categories.

A cardholder who concentrates on dining and supermarket earn would consistently produce more reward value on Gold than Platinum, before factoring fees and credits. The Platinum is for travel-concentrated spenders with strong lounge-access valuation, not for dining-concentrated spenders. We cover this on the Amex Platinum math page.

The two cards are sometimes held together by cardholders with both dining and travel concentrations: Gold captures the food spend at 4x, Platinum captures the airfare at 5x and the lounge access. Combined effective fees and credit stacks need to be evaluated as a portfolio, not card-by-card. The two-card MR portfolio is a common high-end configuration for sophisticated cardholders; whether it produces value depends on the cardholder's actual usage of both stacks.