Methodology
How we score 13 USDT virtual cards, where the data comes from, and what we explicitly don't claim. Public formula, transparent sourcing.
The four-dimension formula
Each card's composite score (0-10) is the weighted sum of four dimensions:
- 35% Fees — USDT topup fee + per-transaction fee + monthly fee + ATM withdrawal + FX spread. Weighted toward topup and per-txn since those compound with usage.
- 25% Reliability — Issuer's official status page history, announcement frequency, community signal (Reddit / Twitter outage reports). Card status downgrades (online → delayed → issue) reduce this dimension for 7 days.
- 20% UX — App quality, KYC friction (basic vs full), supported payment methods (Apple Pay / Google Pay / Alipay / WeChat Pay), card form factor flexibility.
- 20% Compliance — Issuer licensing footprint (MAS, SFC, FCA, FinCEN MSB, MiCA, etc.), regulatory clarity in target markets, history of pause/exit decisions.
Data source ladder
- Issuer official fee page (top priority) — Every card review carries a
feeSourceUrllinking directly to the source. usdtcard.net auto-scrapes this page hourly and republishes any detected change within 24 hours. - Issuer status pages + official announcements — Used for reliability scoring and incident timeline.
- Regulator websites (FATF, MAS, SFC, FinCEN, FCA, European Commission, etc.) — Cited in compliance writeups.
- Public community reports (Reddit, Twitter) — Labeled as "non-verified context" wherever cited.
What we explicitly don't do
- No independent on-chain testing. We do not buy cards, run transactions, or capture transaction hashes. The whole industry map is maintained from public documentation auto-scraped hourly.
- No paid placements or score adjustments from issuers.
- No advance preview of scores to issuers.
- No card recorded without an official fee source page (no source → not indexed).
When scores change
Hourly fee-page check triggers an overnight recompute when a change is detected. Card status downgrade auto-reduces the reliability dimension over the next 7 days. Any composite score change exceeding ±0.5 requires human review before publication.
Why this stance is still valuable
Industry fee data is scattered across dozens of issuer pages with inconsistent framing, currencies, and chain-specific fine print. Comparing them by hand is prohibitively expensive for readers. What we do: structure official data, refresh hourly, and apply editorial judgment for USDT-specific use cases. This is an aggregation + analysis product, not a testing lab. We're upfront about that so you can cite our numbers with confidence.
FAQ
- Q. How is each card scored?
- Each card receives a 0-10 composite of four weighted dimensions: 35% fees (topup + per-transaction + monthly + ATM + FX spread), 25% reliability (status page history + announcements + community signal), 20% UX (app quality + KYC friction + payment method support), 20% compliance (licensed coverage + regulator clarity).
- Q. Where does the data come from?
- Three sources, in priority order: (1) issuer official fee page, linked via feeSourceUrl on every review (top priority); (2) issuer official status pages and announcements, auto-checked hourly; (3) public community reports (Reddit, Twitter), labeled as non-verified context only.
- Q. Do you perform independent on-chain testing?
- No. We do not buy cards, run transactions, or capture transaction hashes. Every review carries a "based on official documentation" badge. Our value is aggregation + comparison + editorial judgment, not test data. We believe public methodology + linkable sources beat a single test run for ongoing reliability.
- Q. Why don't you accept issuer payments?
- usdtcard.net accepts zero issuer payments, runs zero affiliate links, and never previews scores to issuers. This is both our editorial stance and the strongest E-E-A-T signal we can offer. Full disclosure at /disclosure.
- Q. How often do scores change?
- Worker scrapes the official fee page hourly. Any detected change triggers a recompute that night. Card status downgrade (online → delayed → issue) lowers reliability weight over the next 7 days. Any score change exceeding ±0.5 requires human review before publication.