Scoring
How we score your file
Planning scores, not a USCIS grade. Same two-step test officers use. The resume is the claim. The exhibits are the proof. Only proof can mark a box met here.
Cite this method
EB1AGenius scores an EB-1A file as claim versus proof on the ten criteria in 8 C.F.R. § 204.5(h)(3). A box is treated as met for planning at proof of 70 or above, and only when a primary exhibit for that box is on file. The overall number is the whole file, not the average of the bars. USCIS does not use this scale. Method: eb1agenius.com/how-we-score. Sample file: 48/100, 1 of 3 criteria.
We read the resume as the claim
The resume, job titles, and what you say you did are the claim. A strong resume can raise the claim score. It cannot mark a box met on this site.
We read the files as the proof
PDFs, letters, papers, patents, invitations, and press clips are the proof. Each file is mapped to the criterion it actually supports. A published paper is Scholarly Articles. It is not Judging.
A box is met only with a primary exhibit
On this site, proof of 70 or above is treated as met for planning, and only when a primary exhibit for that box is on file. USCIS does not use a 70-point grade. An invitation, an acceptance email, or a manager letter can show a start. Those files cannot open the gate by themselves.
Step 1, then Step 2
Officers use a two-step test (Kazarian). So do we. Step 1 asks whether three of the ten criteria look exhibit-backed, or whether a true one-time major award is on file. Step 2 asks whether the whole record shows sustained acclaim at the top of the field. Three met boxes is not a grant.
The overall number is the whole file
The big number is not the average of the bars. Scholarly Articles at 75 means that one box looks exhibit-backed. If only one of three is met, the overall stays in the building range even if that one bar is high. Final merits (independent recognition over time) still has to be there.
What can actually open a box
Files that are the wrong kind still appear in the evidence matrix. They just cannot push that box to 70 on this site. That is why a review invitation can sit at 60 while a published IEEE paper sits at 75.
Scholarly Articles
Can count as met. A published paper. IEEE Xplore, a DOI, or a proceedings PDF.
Not enough by itself. An acceptance email or a CMT author console. Those show a start, not a published article.
Judging
Can count as met. A reviewer certificate that the reviews were completed, for a journal, conference, or panel.
Not enough by itself. A review invitation with no completed reviews. Hackathon judging is weaker still.
Original Contributions
Can count as met. A granted patent, or independent use outside the employer: a licensee, a lab, or a company that adopted the work.
Not enough by itself. A patent filing, a Google Scholar printout, or a letter from your own company.
Critical Role
Can count as met. Proof the organization is distinguished, and proof the outcome depended on you.
Not enough by itself. A retention letter or a job title on its own.
Awards
Can count as met. A nationally or internationally recognized prize, with who was eligible and who judged it.
Not enough by itself. Employee of the year, team excellence, or a campus award. Those are not a Pulitzer, Oscar, or Olympic medal either.
Published Media
Can count as met. Independent press substantially about you, in professional or major media.
Not enough by itself. Your employer newsroom, a press release you wrote, or a LinkedIn post.
Membership
Can count as met. Admission that required outstanding achievement, judged by experts.
Not enough by itself. Dues-paying or open membership.
High Salary
Can count as met. A pay record and an independent survey for the same specialty and location.
Not enough by itself. An offer letter with no comparison, or a company-only compensation statement.
Exhibitions / commercial success
Can count as met. A juried exhibition, or performing-arts sales and charts, when that is your field.
Not enough by itself. For most software, data, and research files these two boxes are N/A. They do not drag the score.
How the overall number is built
First we score each of the ten criteria. Boxes that do not apply to the field (often artistic exhibitions and performing-arts commercial success for a software file) are marked N/A. Empty and N/A boxes do not drag the average down.
Step 1 is a gate, not a vibe. You need three criteria at 70+ proof, or one true one-time major award (Pulitzer, Oscar, Olympic medal class). A company prize is not that. Until that gate is open, the overall stays in the building range even if one bar is high.
Step 2 is final merits: sustained acclaim, independent recognition, and whether the person sits among a small group at the top of the field. Letters without backup do not carry this.
The overall number mixes those two reads. If Step 1 is not met, more weight stays on the missing boxes. If Step 1 looks arguable, more weight moves to final merits. The public sample is 48/100 with 1 of 3 met on purpose. It is a typical thin file, not a winning showcase.
Bands on this site: Early stage (evidence gaps). Building (focus on 3+ criteria). Promising (strengthen final merits). Strong (filing-ready trajectory). Those labels are ours. They are not a USCIS finding.
What the 30-day list is doing
The next-30-days list starts with the closest box to 70, not the empty one. A judging invitation already on file is an easier lift than starting Awards from zero. That is a collection order, not a promise that the score will move.
What we skip
- Hackathon attendance, schedules, and participation certificates.
- Pages we could not read. A scanned PDF with no text is not scored as an exhibit.
- The same filename printed twice.
- A write-up that says a paper or patent is missing when that file is already on record.
What this is not
Not legal advice. Not a USCIS decision. Not an approval rate. We do not file petitions. Take the report to your own attorney, or use it to decide if you need one yet. Questions: support@eb1agenius.com.
More on EB-1A requirements · Common questions · Scoring tool note.