Rewards are one part of a card decision. This model compares three editable profiles over twelve dated months, keeps ongoing value separate from welcome offers, and makes borrowing assumptions visible; historical examples illustrate their stated inputs.
Follow the complete value bridge
Start with eligible purchases rather than an advertised reward percentage. Allocate every purchase once: a prepaid issuer-portal hotel stay belongs in its eligible portal channel, not also in broad travel. The model earns points or cash according to each profile, applies category and shared limits at their actual reset boundaries, and values points with the conversion you enter. It then adds only the credits you expect to use, subtracts the applicable fee and any included interest estimate, and reports net modeled dollars. Ongoing value excludes the welcome offer; year-one value uses the first-year fee and eligible first-year benefits. Compare that total with the explicitly entered current card, because a positive gross reward can still be worse than keeping an existing card. If two profiles differ by less than one cent, the comparison treats their modeled values as equivalent. This is an arithmetic comparison of the entered assumptions, not approval advice, a product recommendation or a guarantee of realized savings.
Treat redemption and credits as assumptions
A point count is not a cash balance. The selected cents-per-point value converts an earned currency into a modeled dollar amount, while availability, transfer ratios, award pricing, taxes and personal travel plans can change what you receive. Preset valuations remain editable assumptions rather than guaranteed portal or cash returns. Credits need equal care: enter the maximum face amount, the amount you expect to claim and its usable value to you for each reset you expect to use. A benefit that encourages an otherwise unwanted purchase can be worth less than its advertised amount. Monthly, quarterly, semiannual, annual and one-time periods are counted within the dated planning window, including partial periods at either end. When a credited purchase does not earn rewards, sufficient purchases in the matching eligible category are required in each used period. Allocation follows the documented earliest-period convention; the model does not verify a future merchant transaction or predict exactly when a credit will post.
Separate welcome offers from recurring value
A welcome offer can dominate a first-year comparison while saying little about later years. The model therefore requires you to enable the offer, confirm personal eligibility and enter qualifying purchases within its first stated number of planning months. The annual purchase total is not used as a substitute for that window. Qualifying purchases cannot exceed the planned purchases inside it, and the entered threshold must be met before bonus value appears. This simplified monthly window does not establish an issuer deadline expressed in days; use your actual account-opening date and offer terms before relying on the assumption. Some offers are personalized or unavailable in a stable public page, so those presets start with no verified offer and custom fields remain available. Bonuses and first-year credits appear only in the year-one view, while ongoing value keeps continuing benefits separate. Avoid increasing purchases merely to satisfy a target: additional spending, eligibility restrictions and fees can outweigh the modeled reward, and account approval is never assured.
Make borrowing and record limits visible
The paid-in-full assumption appears next to results because a reward comparison can be misleading when financing costs are omitted. If you carry a balance, each profile can include a simple estimate based on its average carried balance, APR and number of months; the formula is balance times APR times months divided by twelve. This is not a daily-compounding amortization model, a statement forecast or a determination of grace-period eligibility. Payment-based reward components separately use the fraction of purchases paid during the modeled period. The dated U.S. product records cover selected core rates, fees, credits and publicly confirmed offers, with merchant exclusions and unmodeled benefits disclosed as limitations. They remain available as editable snapshots if a product page changes or cannot be reached. Your comparisons stay in the browser, and Remember is optional. CSV retains full-precision canonical assumptions and both result views; a share link contains the comparison you explicitly review, so check its contents before sending it to someone else.