What I learned verifying 73 Indian credit cards against issuer terms

I started CardMasters because the headline reward rate was rarely the number that decided whether a card was good.

After checking all 73 active cards in the catalogue against issuer pages and terms, four patterns kept repeating:

• A monthly cap often matters more than the advertised percentage.

• Merchant-routed rewards should not be stretched across generic online spending.

• Fees, redemption costs and exclusions can erase a high earn rate.

• Affiliate commission has to stay outside the ranking model.

The product now runs every public ranking and the logged-in wallet recommendation through the same deterministic scoring engine. Each computed card needs a dated issuer-source check; approval remains an estimate, never a bank decision.

I also published the evidence surface so the assumptions can be challenged:

https://www.cardmasters.in/credit-card-changes?utm_source=indiehackers&utm_medium=organic_social&utm_campaign=issuer_terms_lessons_202608&utm_content=founder_post

For founders building trust-sensitive products: what would make you trust a recommendation more—the exact calculation, the original source, or visible outcome feedback from other users? I would value blunt feedback on what is still missing.