Original contributions of major significance: usually the criterion the whole case turns on

If you only have time to get one criterion right, this is probably it — it's the one AAO decisions come back to most often, and it's the one that's easiest to argue in general terms without actually meeting the standard. The standard has two separate parts, and both have to be there: 1. Original — a specific, named contribution: a method, a finding, a product, a technique. Not "my work in X" — the actual thing you contributed. 2. Major significance — independent evidence that the contribution has already had a documented impact on the field, not that it's promising or could be significant later. The most common gap isn't a missing contribution — it's evidence that describes the petitioner generally ("a talented engineer," "a leading voice in the field") instead of identifying the specific contribution and explaining, mechanically, why it mattered. A reference letter that praises you is not the same as a reference letter that explains what changed in the field because of a specific piece of work. Citation counts alone have the same problem. A number with no explanation of what the citations show — are other researchers building on the method, citing it as background, or just listing it — doesn't establish significance by itself.