Picking a Buildathon Project Isn't a Product Decision: It's a Positioning One

Four exciting ideas sat on the table going into SheBuilds, Lovable's buildathon for women builders. All four felt viable. That's exactly the situation where picking by gut feeling is the most tempting and the least reliable, so I didn't.

Instead I used AI to actually research what wins a competition like this, rather than guessing from instinct alone. Specifically: Season 1 submission patterns, and the judging criteria as written, not as assumed.

What came back reframed the decision more than I expected. B2C was the single most crowded category, meaning whatever idea I entered there would be competing against the largest pool of similar submissions. Social Impact, by contrast, drew far less competition. And the single heaviest-weighted judging factor, more than polish, more than scope, was whether you could answer one specific question convincingly in a two-minute demo: why you?

That reframe pointed directly at Switchpath, a career portfolio builder that matches resumes to job postings and generates project briefs to prove competency, over the other three ideas that had felt more exciting going in. The reason was simple once the criteria were clear: two years spent teaching AI skills to career switchers was exactly the kind of "why you" answer that question was built to reward. None of the other three ideas had that same personal, provable answer sitting behind them.

The research-first prompt

Before picking which idea to build for [competition/buildathon name]:
research past winners and submission patterns. What categories are most
and least crowded? What does the judging criteria actually weight most
heavily? Then help me honestly match my strongest idea against those
criteria — not which idea I'm most excited about.

Here's the actual lesson, and it's not "AI picked my project for me," which would be the wrong takeaway. AI didn't make the decision. It helped me understand what criteria actually mattered in this specific competition, so I could match my own ideas honestly against them instead of defaulting to whichever one felt most exciting in the moment I was choosing.

Picking a buildathon project, or really any project entering any competitive selection process, isn't only a product decision. It's a positioning decision, and positioning depends on knowing what's actually being judged before you commit.

If you're choosing between multiple project ideas for something with real judging criteria attached, don't start with excitement. Start with research: what's actually rewarded, what's overcrowded, and then ask yourself honestly, the way that question forced me to: where do you have the most unfair advantage?