Your outcome: A weekly review routine and one specific improvement to test.
1. Work from conversations, not guesses
Review new replies in Inbox and questions in Tasks. A contact that replied can enter your CRM pipeline. Keep Flow for the pre-CRM outreach steps. Assign a clear next action so a promising enquiry does not disappear.
2. Compare the right numbers
Review the campaign statistics and search data you actually have. Compare leads with views for website campaigns, and genuine replies with delivered outreach. Small samples can mislead: record the period and context before concluding a message or page is better.
20 leads from 1,000 campaign views is 2%. If you change the offer, compare a similar audience and period; do not assume the change alone caused the result.
3. Turn one discovery into a better system
Look for repeated questions, poor-fit leads or delayed follow-ups. Approve a verified learning into Knowledge, refine an audience or change one offer. Keep a note of what changed and check the next comparable period. The course does not automatically optimise campaigns or send messages.
Quick knowledge check
How does an AI-proposed learning become a fact the AI can use?