Screenshots use a demonstration workspace. Your data, plan, and activation status may differ.
Choose a period of **7, 14, 30, or 90 days**. Narrow by page, source, or desktop/mobile where those values are available. Review the active filters before comparing numbers. ## Conversion Follow the funnel from visitors and displayed openings to chat opens, conversations, contacts, and recorded meetings. The rate between steps uses the previous step as its denominator. Headline comparisons use the equally long preceding period. An empty rate or unavailable comparison means the denominator or earlier evidence is missing. It does not mean Darwin converted zero percent of an audience that was never measured. ## Pages & sources Compare page and acquisition-source rows using both volume and outcomes. A page with two visitors and one meeting is different from a page with hundreds of visits; do not choose an opening from rate alone. Use page-level results to identify where conversations stall, where your knowledge is incomplete, or where an experiment deserves closer review. ## Conversation quality Quality reporting summarizes analyzed conversations, intent, buying stage, qualification, and observed failure categories. The sample-size message matters: a small number of graded conversations cannot establish a reliable trend. Open the suggested transcripts to understand the reason behind a score. AI-generated classifications should guide review rather than replace it. ## Test impact Review promoted openings, their recorded result at promotion, and subsequent exposure. A recorded lift describes the experiment’s expected-value comparison, not a guaranteed causal increase in every later meeting.