N8N WORKFLOW SETUP
Add Facebook Insights reporting to an n8n workflow
Connect Adzviser to an AI Agent, test one request, and only then add the schedule, alert, or delivery step.
Connect Facebook Insights to Adzviser
In Adzviser, sign in to Facebook Insights and allow Adzviser to read its reporting data.
Configure the MCP Client Tool
Add n8n's MCP Client Tool to an AI Agent, enter Adzviser's server address and credentials, and test the tool before activating the workflow.
https://mcp.adzviser.com/httpTest before activating
Run the workflow manually with one Facebook Insights request. Check the selected account, dates, and output before turning on its trigger.
AUTOMATION BLUEPRINT
Build a repeatable Facebook Insights reporting workflow
Keep the automation easy to audit: one trigger, one clearly scoped analysis, and one useful destination for the result.
Decide when it runs
Use a schedule, webhook, or manual trigger and calculate a complete reporting period before requesting data.
See a related Facebook Insights promptGive the AI Agent a stable instruction
Fix the account-selection rule, metrics, breakdowns, comparison, and output format so each run is consistent.
See a related Facebook Insights promptSend a result people can use
Route the takeaway and evidence table to email, Slack, a database, or the next approved step in your workflow.
See a related Facebook Insights promptN8N AI AGENT INSTRUCTION
Use a structured Facebook Insights instruction in every run
A predictable output makes the automation easier to test, route, and monitor.
Understand what drives follower growth.
First show my connected Facebook Insights accounts and ask me which one to use. Then complete this analysis: Analyze Facebook Insights follower growth for the workspace I choose during the last 30 complete days. Show: Daily New Likes, Daily Unlikes, Net Growth, and growth rate. Correlate growth spikes with specific posts or events. What content drives the most new followers? Recommend growth acceleration strategies. Return valid JSON with keys summary, evidence, alerts, and next_steps so the following n8n node can route the result reliably.
