"The 3 things AOs actually care about"
- Narrative coherence — the project must connect to who the student is and what they want to study
- Intellectual depth — did the student go beyond the tutorial? Can they explain why it works, not just that it works?
- Measurable impact — "I built it" is table stakes. "200 people used it and their grades improved" is interesting
Acknowleges: Pure research projects (ISEF, arXiv) carry real prestige weight that no app can match. Students who can do rigorous original research should be encouraged to do so — but it's not the only path.
"What I'd hire for: depth over breadth"
- Production-grade code matters more than novelty — did they write tests? Handle errors? Use version control properly?
- System design thinking: can the student explain tradeoffs? Why Postgres not Mongo? Why REST vs. GraphQL?
- Open source contributions to real projects show you can work in other people's codebases and communities
Acknowleges: Product thinking (building for real users, handling feedback) is genuinely hard to fake and shows a level of maturity that side projects rarely demonstrate.
"Users are the only proof that matters"
- Any student can build a project. Getting 100 people to use it consistently? That's the real signal
- Revenue — even $50/month — proves someone values what you built enough to pay
- Building something for your immediate community (school, neighborhood) is a feature, not a bug
Acknowleges: Technical depth is the one thing I can't teach in a weekend. A student who genuinely understands systems programming or algorithms has a durable advantage that product intuition alone can't replace.
"Original contribution is non-negotiable"
- The gold standard is contributing new knowledge — even at HS level, a well-executed ISEF project or published paper changes the conversation
- Methodology matters: hypothesis, data collection, controls, analysis, peer review
- Work that could be a class assignment won't fool anyone; work that extends beyond the curriculum is what we're looking for
Acknowleges: Publishing is not the only path and is increasingly not the best path for most students. A well-documented, thoughtful app with real users can demonstrate equally rigorous thinking in a more accessible format.
"Be honest about what you can actually finish"
- Most HS students overestimate what they can build in 6-18 months. Scope control is a survival skill
- The best project is one you actually finish and can talk about passionately in an interview
- Accessibility and underrepresented communities are a differentiator that AOs are actively looking for
Acknowleges: Hardware projects look incredibly impressive on paper but are prone to catastrophic scope creep. The engineers on my team who shipped fastest all started with software-only projects.