Zach Roseman’s Post

Why Draftboard v1 Didn’t Work: A Post-Mortem First up: the long feedback loop. Hiring is notoriously time-consuming. Even after crafting and posting a job description—a process in itself—there’s a lengthy path from sourcing to screening to scheduling and, finally, onboarding. We’re talking months, not weeks. For an early-stage startup like ours, speed is everything. To iterate and improve the product, user experience, and business model, we need rapid cycles of feedback. Draftboard v1 didn’t allow for that. While we could quickly measure top-of-funnel metrics—new scouts, applicants, referrals, first interviews—it took far too long to gather data on what happened at the bottom of the funnel, such as offers made, offers accepted, and candidate retention. The result? We were only able to make adjustments to the top of the funnel, while critical insights from the hiring outcomes were too delayed to act on effectively. Without a fast feedback loop, feature prioritization became guesswork. Any action taken to address bottom-funnel issues would take months to show results—by then, we were often focused elsewhere. Ultimately, this wasn’t sustainable. We reached a point where product improvements aimed at bottom-funnel challenges had low confidence for success and would take far too long to validate. In a startup world driven by limited time and resources, this approach just doesn’t work. It became clear we needed to fundamentally reassess whether our core hypothesis was still viable—and, as it turns out, it wasn’t. Stay tuned for tomorrow’s post: The Leaky Bucket.