KEY TAKEAWAYS
- Cost per request is misleading; cost per classified card is the real metric.
- Low classification quality raises effective cost through retries.
- Deduplicate lists and retry only errors to control spend.
A cheap per-request price is not the same as a cheap result. Every error forces a retry, and retries spend balance twice for one eventual answer. The honest metric is cost per classified card.
Our numbers
- Total spent on checks in this window: $21.78
- Paid classified results: 278
- Effective cost per classified card: $0.0783
| Gateway | Paid classified | Spent | Cost / classified |
|---|---|---|---|
| Fast Checker | 84 | $1.66 | $0.0198 |
| Pre-Auth Gateway 2 | 61 | $9.54 | $0.1564 |
| Gate3 Charged | 42 | $1.68 | $0.0400 |
| Pre-Auth Gateway | 91 | $8.90 | $0.0978 |
Why the effective cost differs from the sticker price
Gateways bill per classified result, so the sticker price per check only applies when the check classifies. Providers with lower classification quality push effective cost up through retries; providers with stable answers push it down. That is why we compare gateways on classified percentage and p95 latency as well as price.
How to keep effective cost low
- Deduplicate lists before running them — duplicate cards are duplicate spend.
- Retry only errors, never dead cards.
- Use a fallback gateway so transient errors still classify on the first pass.
- Track cost per classified card monthly, not cost per request.
More in the bulk checking workflow and error-rate engineering.
Methodology
Totals come from platform billing records over the last 90 days; only classified (live/dead) outcomes incur charges. Free promotional checks (new-user free checks) are excluded from these cost figures. No customer data is used or published.