The founders who get a clean O-1A approval did not have a better story. They had a record. And they started building it long before any lawyer touched a Form I-129.
Here is the uncomfortable truth most agents will not tell you: if you are reading the O-1A criteria for the first time and you plan to file in two months, you have probably already lost the strongest version of your case. Not because you are not extraordinary, you may well be, but because the evidence that wins this petition takes time to manufacture honestly. A judging invitation you accept this week becomes a credential next quarter. Press you earn in March anchors a petition you file in December.
So we tell most founders the same thing: do not file yet. Spend the next 12 to 18 months engineering the record. Below is how.
Why "not yet" is the right answer
The O-1A is for someone with extraordinary ability in the sciences, education, business, or athletics, the small percentage at the very top of the field. You qualify one of two ways. Either you hold a single, internationally recognised honour (a Nobel, an Olympic medal) and you are done, or, the route nearly every founder takes, you satisfy at least three of eight evidentiary criteria.
The eight are: national or international awards; membership in associations requiring outstanding achievement; published material about you in major media; authorship of scholarly articles; original contributions of major significance; judging the work of others; a high salary relative to your field; and a critical or essential role at a distinguished organisation.
Here is the part founders miss. In 2026, USCIS does not stop at counting boxes. After you clear three criteria, the officer runs a second pass, a totality or final-merits review, asking whether the whole record actually shows someone at the top of the field. A single high-impact paper with hundreds of independent citations beats five papers nobody cites. Strong evidence across three criteria beats thin evidence smeared across seven. You cannot fake depth in six weeks. You can build it in a year.
And several of these criteria are not things you have. They are things you do. You do not stumble into a judging seat or a TechCrunch feature. You go and get them. That is the whole game.
The campaign, criterion by criterion
Judging, the most underused, and the easiest to manufacture honestly
Judging is the criterion most founders ignore and the one you have the most control over. You evaluate the work of others in your field. USCIS in 2026 reads this broadly: sitting on a pitch competition jury, reviewing grant applications, serving on a technical conference programme committee, assessing startup applications for an established accelerator, even reviewing pull requests for a major open-source project at a senior level.
This is a campaign you can start tomorrow. Email three accelerators and offer to mentor or judge a demo day. Volunteer for the review panel at a sector conference. Get on the screening committee for a startup award. Each one produces a paper trail, an invitation email, a confirmation, your name on a published jury list. Bank six to eight of these over a year and you do not just meet the criterion, you meet it convincingly.
Published material, earn it deliberately, do not wait for it
Tier-1 press about you and your work is one of the cleanest criteria, and the one founders most passively hope will "happen." It will not. You engineer it. USCIS now explicitly counts digital publications, expert podcast appearances, and serious online outlets, not only print.
The play is a sustained record, not one lucky hit. Brief a PR person early. Tie coverage to real events: a funding round, a product launch, a piece of original data only you have. Pitch yourself as the expert source for the journalist already writing the trend piece. The goal is a steady drumbeat of named coverage in recognised outlets across the build window, so that by filing you can show a record rather than a coincidence.
Scholarly articles and original contributions, publish, patent, get cited
You do not need a PhD. You need authorship and impact. Write for a respected industry journal. Co-author a paper with an academic collaborator. File a patent. Open-source something the field actually adopts. The strongest evidence here is independent uptake, citations, adoption, other people building on your work, because that is what the final-merits review rewards. Start the publication and patent timelines early; both move slowly.
Membership and selective programmes
Membership must require outstanding achievement judged by recognised experts, not a paid directory. Acceptance into a genuinely selective accelerator (Y Combinator, Techstars) can support both the membership and the awards angle, provided you document the acceptance rate. Apply on your own timeline, not under visa pressure.
Critical role and high salary
As a founder, the critical-role criterion is yours almost by definition, the work shifts to proving your company is distinguished (funding, traction, named investors). High salary is harder pre-profit, but a market-benchmarked founder package, properly documented, can contribute.
Where founders waste the most effort
One trap above all: treating a VC round as an award. It is not. USCIS has been explicit in 2026, raising capital, on its own, does not satisfy the awards criterion. Investors are buying equity, not certifying you as among the best in your field. A funding announcement is still enormously useful, but as fuel for press and as evidence of a distinguished organisation, not as a prize. Pair it with a competitively judged award if you want awards on the board.
The second trap is pay-to-play. Awards bought through entry fees, vanity memberships, and "top 40" lists with no real selection are not just weak, in 2026 they actively signal a thin case. One competitively judged honour outweighs ten you paid to receive.
| Move | Counts for the founder? | Why |
|---|---|---|
| VC funding round, on its own | No (not as an award) | Equity purchase, not a competitive honour, but powerful for press and critical-role evidence |
| Competitively judged startup award | Yes | Recognised selection by experts beyond your company or region |
| Selective accelerator (YC, Techstars) | Yes | Documented low acceptance rate; supports membership and awards |
| Paid "top founders" list | No | No genuine competitive selection; signals a weak case in 2026 |
| Judging a real pitch competition or grant panel | Yes | Direct evidence of evaluating others' work in the field |
A 12–18 month build plan
Below is the campaign as a timeline. The exact months matter less than the sequence: slow-moving evidence first, the funding-and-press crescendo timed to land a few months before you file.
| Window | What you build | Criteria served |
|---|---|---|
| Months 1–3 | Map your gaps with an advisor. Start patent/publication timelines. Apply for 2–3 judging or review seats. Brief a PR person. | Judging, scholarly articles, original contributions |
| Months 3–6 | First judging seats confirmed. First named press placements. Apply to one selective accelerator or merit-based body. | Judging, published material, membership |
| Months 6–9 | Publish your article or paper. Patent filed. Line up 5–6 recommenders, a mix of independent experts and people who know the work directly. | Scholarly articles, original contributions, recommender base |
| Months 9–12 | Time the funding announcement here, use it to drive a press cluster, not as an "award." Add 2–3 more judging credentials. Track citations and adoption. | Published material, critical role, original contributions |
| Months 12–15 | Collect and refine recommendation letters. Document accelerator acceptance rates and award selection processes. Benchmark your salary. | Membership, awards, high salary, critical role |
| Months 15–18 | Assemble the petition. Run the final-merits stress test, does the whole record read as top-of-field? File. | All eight, totality review |
Recommenders: start early, choose carefully
Strong O-1 letters are the slowest item to produce and the most often rushed. You want a blend: independent experts who can speak to your standing in the field generally, and people with direct knowledge of your work specifically. Approach them six to nine months out, not six weeks. A letter drafted under deadline pressure reads like one. A letter from someone who has watched your record build for a year reads like the truth, because it is.
How SaathiVisa thinks about this
We would rather tell a founder "come back in fourteen months" with a build plan than file a thin petition now and absorb a request for evidence, or a refusal that follows you. The clients who get clean, fast O-1A approvals are the ones who treated the visa as a 12-to-18-month project, not a paperwork sprint. If you are early, that is good news. It means the strongest version of your case is still available to you. We help you go and build it.
FAQ
Can I really start building O-1 evidence before I have a lawyer or a sponsor?
Yes, and you should. Judging seats, press, publications, and patents are credentials you accumulate regardless of who eventually files. The earlier you start, the stronger and more credible the record. The petition itself comes at the end.
Does raising a venture round get me the O-1A?
No. As of 2026, USCIS does not treat a funding round as an award, investors are buying equity, not certifying you as top of your field. The round is still valuable: it powers press coverage and helps prove your company is a distinguished organisation. Pair it with a competitively judged award for the awards criterion.
Is it better to meet more criteria or fewer with stronger evidence?
Fewer, stronger. After you clear three criteria, USCIS runs a final-merits review of the whole record. Deep, well-documented evidence across three criteria consistently beats thin evidence spread across seven. Build for depth.