A “5% cashback” card can be worse than a “2 points per ₹100” card—or much better—depending on category caps, exclusions, redemption value, fees and your own spending. The comparison must use your last 6–12 months of normal spending, not the issuer’s best-case marketing example.
Build a category-by-category value sheet
| Spend category | Annual spend | Cashback/points earned | Caps/exclusions | Net value |
|---|---|---|---|---|
| Groceries | Actual spend | Use live issuer rate | Monthly/annual cap | Redeemable ₹ value |
| Fuel | Actual spend | Check surcharge/points rules | Often treated differently | Net after fees |
| Utilities/rent/wallet | Actual spend | Check eligibility | Common exclusion area | Only count eligible value |
| Travel/dining | Actual spend | Category multiplier | Merchant-code dependence | Real redeemed value |
Use realised value for points
For reward points, calculate (cash price you would genuinely pay − unavoidable redemption fees/taxes) ÷ points used. A point is not worth the highest value ever achieved by a travel blogger. If you mostly redeem ₹500 vouchers, use that value.
Subtract annual cost and induced spending
Net annual value = cashback actually received + reward value actually redeemed + benefits you would otherwise buy − annual fee − applicable taxes/charges − extra spending caused by fee-waiver or milestone thresholds. A card that makes you spend ₹20,000 unnecessarily to “unlock” a ₹1,000 benefit has negative incremental value.
Run a downside case
Reward programmes change. Recalculate value if a major category is capped, a transfer ratio worsens, or a lounge benefit becomes spend-linked. Do not choose a card whose entire value depends on one fragile perk.
Issuer-specific reward rules control the economics, so verify the current MITC and rewards pages directly at the issuer. RBI’s credit-card FAQ is the baseline for regulated card conduct, not a guarantee of any commercial reward rate.
Decision rule: choose the card that produces the highest conservative net value from spending you would make anyway. If the calculation needs aspirational travel, perfect merchant coding and maximum caps every month, it is not a realistic comparison.