Original contribution: what it is

The hardest criterion of the ten. Over the site base of 1,242 AAO appeal decisions it is counted in 16% of the cases that claim it, and in an external sample of USCIS requests in 4% (7 of 161); across the 137 notices reviewed for the EB-1A page the officer engaged with it in detail 48 times and accepted it zero times. The reason is one: almost everyone proves only the first half - what they made - while the criterion asks you to show the reaction of the field. Those who do pass it clear the second step (final merits) noticeably more often than average: it is exactly the influence the officer assesses there.

Two steps and the boundary. Policy Manual: first USCIS determines whether there is an original contribution, then whether it is of major significance to the field; funding, a patent or a publication show originality but do not by themselves prove significance. Denials constantly cite Visinscaia v. Beers: you need substantial influence beyond your employer, clients and customers. Doing good work for a company and building a product inside it is the expected behavior of a professional, not a contribution to the field.

What counts - in descending weight. Traces of adoption: license agreements and contracts with revenue, adoption records from organizations that are not your clients, regulations and SOPs of other organizations with your method, inclusion in an adopted industry standard, verifiable metrics from an official source. Independent recognition: citations with context (how many you have versus the average in your sub-field - officers demand a comparison, not a number), independent reviews and case studies, industry publications on the application. Interpretation: 2-4 expert letters per contribution, where every statement points to a specific exhibit and the author explains how they know the work.

What does not count. A patent without use (the significance of the innovation is not evaluated when a patent is granted), a publication without traces of influence, moderate citation without context, authorship of a course or a book by itself, claims of adoption without documents, LinkedIn and Wikipedia (zero weight under Badasa v. Mukasey), letters only from personal acquaintances or without the physical address of the author, templated letters with an identical structure - in 2025-2026 officers write outright that such letters look AI-written. Achievements only from the last year may pass formally but read at the second step as preparation for the petition.

How to build it and how to read the RFE. Pick 1-3 contributions, not ten - weak evidence provokes a request about exactly that; for each, start from the "adoption" layer, not from the artifact: if there are no traces beyond your own company, the criterion is probably not yours, and the critical role or high salary is worth a look. The RFE phrase "Second, USCIS determines whether the original contributions are of major significance to the field" is the standard second-step template: the answer is built on independent proof of influence, not on new descriptions of the work; "highly cited relative to others" means a comparison is required, not an absolute number.

In O-1A (item 5 of eight) and EB-1A (8 CFR 204.5(h)(3)(v)) the wording of the criterion is the same, so the breakdown above applies to both visas. EB-2 NIW has no criteria list, but the same evidence of influence works for Dhanasar prong 2 - that you specifically can advance the endeavor. Citation and h-index benchmarks are on the evidence page; officer wording on this criterion is on the RFE page.