Built because the spreadsheet stopped working
Six cards, a tracking spreadsheet that was always too generous, and a growing suspicion the whole arrangement earned less than it looked like on paper. This is the same question, answered properly.
When the data is ambiguous, assume less.
Rewards content has an incentive problem — sites are paid when you open a card, and the numbers drift upward accordingly. The rule here runs the other way.
Lower of two rates
A card publishing 3x and 2x for the same category is modeled at 2x. The higher tier is usually condition-bound.
Spend-gated perks excluded
A "$4,750 in credits" headline that needs $250,000 of prior-year spend is shown separately and counted at zero.
No assumed bonuses
Unannounced rotating categories score at the base rate. Cards with no published base rate score 0x and are never recommended.
The conflict is handled in the architecture.
This site may earn money when someone opens a card through it. So the code that ranks cards has no access to which cards pay — it's a separate file the ranking function never reads, enforced by an automated test that fails the build if that wall is breached.
Three things being built next.
Approval odds
Next Card currently ranks on value alone — it should also weigh 5/24, income and issuer rules before suggesting anything.
Card-linked offers
No issuer publishes an API for Amex or Chase Offers — which is exactly why they go unclaimed, and exactly why it's worth solving.
Mobile
The version that tells you which card to pull out while you're standing at the register.
Credit Upside is built and run by one person — me, Eric — as Credit Upside LLC in New Jersey. Not a content farm, not a lead-gen shell: a real tool I wanted for myself and now maintain in the open. If a number looks wrong, or you just want to talk, email me directly at admin@creditupside.com — it reaches me, not a queue.
— Eric, founder