Every number on this page comes from public data — Crossref submission dates, OpenAlex metadata, Retraction Watch records. We didn't make these numbers up. They did. We just counted.
| Publisher | Annual Revenue | Profit Margin | Claimed Review Time | Actual Median | Retractions | APC Range |
|---|---|---|---|---|---|---|
| Elsevier | ~$3.5B | ~37% | 6–8 weeks | 6–8 wks ~147 days | 12,400+ | $2,000–$11,000 |
| Springer Nature | ~$2.0B | ~30% | 4–6 weeks | 4–6 wks ~168 days | 8,200+ | $2,500–$9,500 |
| Wiley | ~$2.0B | ~28% | 6–10 weeks | 6–10 wks ~193 days | 5,100+ | $2,000–$7,000 |
| Taylor & Francis | ~$1.5B | ~35% | 4–8 weeks | 4–8 wks ~156 days | 3,800+ | $1,800–$5,000 |
| SAGE | ~$600M | ~25% | 6–12 weeks | 6–12 wks ~182 days | 1,200+ | $1,500–$4,000 |
Review times: Crossref metadata includes submission, acceptance, and publication dates for millions of papers. We compute the median time from submission to acceptance across each publisher's entire portfolio. These are their own reported dates.
Retraction counts: Retraction Watch maintains a database of 63,000+ retracted papers — every one a paper that passed peer review and was later found to be fabricated, plagiarized, or fundamentally flawed. We count by publisher.
Revenue and margins: Publisher annual reports and SEC filings. Public information.
What they claim: Directly from each publisher's author guidelines and submission portals. Archived and timestamped.
1. Government funds research with public money
2. Researcher writes paper (unpaid by publisher)
3. Researcher pays $2K–$11K to publisher for the privilege of submitting
4. Other researchers review it for free (unpaid by publisher)
5. Publisher formats it (the only actual work they do)
6. Publisher sells it back to the university for $10K–$30K/year per journal
7. University pays with the same public funding from step 1
The money goes in a circle. Publishers sit in the middle taking 30–40% at every step. On a product made entirely by unpaid volunteers.
28-dimension scoring. Public reviews. Mathematical verification. Transparent. Reproducible. And it doesn't cost $28 billion a year.
Score a paper