Attended my first Claude hackathon and learned that building a good product and winning a hackathon are two very different things

I recently attended the Claude OC Impact Lab hackathon, and it completely changed how I see hackathons as a software engineer. I thought hackathons were like science fairs, where solutions are judged on practicality, correctness, and how well they solve the problem. Turns out, not really. Around 29 of 31 teams chose the same challenge: building a referral intake system to help coordinators review 15–35 pages of medical documents and quickly determine whether patients qualify for recovery care. The rubric emphasized day-one impact, usability, and demo clarity. But from what I observed, organizers used Claude to analyze the submitted codebases, with less than two hours to judge roughly 35 teams. I'm skeptical how thoroughly they could evaluate actual functionality and correctness. What surprised me most was that two winning team's apparent differentiating feature was adding Spanish language support to their website . Don't get me wrong, multilingual accessibility is great, and I do not think my product will win even with it. But the intended users were referral coordinators, not patients . How does translating the interface improve eligibility decisions, processing speed, or accuracy? Especially when adding multilingual support with AI is relatively trivial. My biggest takeaways: Perceived functionality matters more than actual functionality. A flashy demo with lots of features can beat a smaller, more reliable solution. Easy but impressive-looking features have huge ROI. Meanwhile, error handling, validation, and backend reliability are practically invisible. Presentation and networking matter. Talking to organizers and understanding what impresses them can be more valuable than technical complexity. Choose teammates carefully. Spending half a one-day hackathon in meetings instead of building isn't ideal. The irony is that winning a hackathon can incentivize the opposite of good software engineering : more features over reliability, appearance over functionality, and presentation over correctness. I understand judges have limited time, and I'm not claiming the winners didn't deserve it. I may have missed other strengths of their projects. But I went in trying to build something genuinely useful and left realizing that hackathons are largely about selling an idea. Is this normal for hackathons, especially ones judged by non-technical people?