AI Stack for Meta Ads B2B

Meta Ads is rarely the primary B2B channel — but for companies already running LinkedIn at scale, it's often the most cost-effective retargeting surface for accounts already in your funnel. This stack makes the Meta investment measurable. Pull campaign performance data from the Meta Ads API, cross-reference against CRM lead quality data to identify which Meta audiences are generating pipeline-qualified leads, and run the combined dataset through a weekly performance summary prompt. The output: a structured view of which Meta campaigns are worth sustaining and which are burning on volume without conversion quality.

The Stack

The Prompt

This stack is built around the Weekly Paid Media Summary Prompt. Here's the abbreviated version — the full prompt with all variables and usage notes is on its own page.

You are a B2B paid media analyst evaluating Meta Ads performance for pipeline impact.

Review the campaign performance and CRM lead quality data provided below.
For each active Meta campaign, report: spend, CPL, lead volume, and CRM-qualified pipeline generated.
Identify which Meta audiences are converting to pipeline versus generating volume without qualification.
Flag any campaigns with CPL significantly above the target threshold or zero pipeline contribution.
End with 3 specific optimization recommendations: audiences to suppress, custom audiences to build, and campaigns to scale or pause.

The Workflow

1. Export Meta Ads campaign data

   Pull the past 30 days of campaign performance from the Meta Ads API or Ads Manager — spend, impressions, CPL, and lead form submissions by campaign and audience.

2. Pull CRM quality data for Meta leads

   Export Meta-sourced leads from Salesforce — including ICP score, MQL status, SAO rate, and any pipeline created. Map leads back to Meta campaign using UTM data captured at form submission.

3. Build custom audiences from CRM segments

   In Meta Ads Manager, create or update custom audiences from your highest-quality CRM segments — ICP-matched accounts, open opportunities, and recent closed-won lookalikes.

4. Run the performance summary prompt

   Paste the combined Meta performance and CRM quality data into the Weekly Paid Media Summary prompt. Claude identifies which campaigns are generating qualified pipeline and which are just generating volume.

5. Restructure and report

   Restructure Meta campaigns based on Claude's recommendations — pause low-quality audiences, expand pipeline-generating ones, test new custom audience segments. Log decisions in Sheets and post summary to Slack.

What This Replaces

• Meta Ads optimization based solely on platform-reported CPL without CRM quality validation
• Retargeting audiences built once and never refreshed with updated CRM segments
• Meta budget decisions made in isolation from the broader channel mix and pipeline contribution data

Related Stacks

• AI Stack for Paid Media Reporting →
• AI Stack for Ad Creative Testing →
• AI Stack for B2B Paid Media Attribution →