IntentBridge: Workflow-Driven Onboarding for Low-Intent Acquisition Channels

30 days later. A follow up from my Side Project hitting 500 users.

30 days later. A follow up from my Side Project hitting 500 users.

Your funnel theory matches my data almost exactly, from a much smaller project: my first real user came from an AI answer (ChatGPT cited one of my guide pages) and converted the same day, while story-driven traffic arrives curious rather than with a problem, browses, and leaves. The uncomfortable conclusion: the channel doesn't just bring users, it brings intent, and no onboarding fix fully compensates for intent that was never there. So before spending a month closing the signed-up-to-wired-in gap for the Reddit cohort, I'd double down on the channel that already sends purposeful people — more marketplace surface area, more pages answering the exact questions those users ask. On the public workspaces question: don't hunt for the "generally interesting" dataset — look at what your power users are already loading into their private graphs by hand. If three of them have independently built the same corpus (a doc set, a standard, a body of regulations), that's your first public workspace, pre-validated. And I'd bias toward data that changes — rules, pricing, APIs. A living graph justifies a subscription; a static corpus is a one-time download someone will clone. Also, moving hosts and giving every sub a free month after the outage was the right call — that's exactly the kind of thing the users who really get into it remember.