LeadForge reads what your published content actually did — which pieces produced leads, which personas responded — and turns it into a single specific instruction for this week. You approve it or ignore it; either way you never start from a blank page.
Every insight is derived from your attribution data, not from a general model of "what works on LinkedIn". If a format or a topic outperformed for you, that’s what shows up here — and if there isn’t enough data yet to say, LeadForge says that instead of inventing a trend.
The recommendation arrives with the channel, the persona and the goal already attached — approve it and it becomes a brief, drafted in your voice, waiting for review. Park it and LeadForge stops suggesting it. Nothing moves on its own.
What you publish gets tracked, and what it produced changes the recommendation you see next week.
Recommendations are ranked by what produced qualified conversations, not by reach or engagement.
Query your own attribution in chat when you want the reasoning behind a recommendation.