wallacedobbs428/ads-report

Ad Report — Performance Dashboard

Pull performance report for: $ARGUMENTS (or all active campaigns if no vertical specified)

Context

The Call Taker: AI receptionist SaaS. CPL benchmarks:

• Towing: $12 target / $24 kill
• Locksmith: $15 / $30
• HVAC: $18 / $36
• Plumbing: $20 / $40
• Roofing: $22 / $44
• Dental: $35 / $70

Step 1: Load Active Campaigns

If a vertical was specified, read:

~/thecalltaker-ops/ads/active/{vertical}-campaign.json

If no vertical, glob all: ~/thecalltaker-ops/ads/active/*-campaign.json

If no campaigns found: "No active campaigns. Run /ads-launch {vertical} first."

Step 2: Check for API Tokens

echo "META_ACCESS_TOKEN: ${META_ACCESS_TOKEN:+SET}"

If tokens exist, pull live data from Meta API:

curl -G "https://graph.facebook.com/v21.0/${CAMPAIGN_ID}/insights" \
  -H "Authorization: Bearer ${META_ACCESS_TOKEN}" \
  -d "fields=spend,impressions,clicks,cpc,cpm,actions,cost_per_action_type" \
  -d "date_preset=last_7d"

If no tokens, generate report from campaign config with "PENDING_LAUNCH" status.

Step 3: Extract Metrics

For each campaign:

• Spend: Total $ spent
• Impressions: Times shown
• Clicks: Link clicks
• CTR: Click-through rate
• Leads: From actions where action_type = "lead"
• CPL: Cost per lead (spend / leads)

Per-ad breakdown:

curl -G "https://graph.facebook.com/v21.0/${ADSET_ID}/insights" \
  -H "Authorization: Bearer ${META_ACCESS_TOKEN}" \
  -d "fields=spend,impressions,actions,cost_per_action_type" \
  -d "level=ad" \
  -d "date_preset=last_7d"

Step 4: Generate Recommendations

For each ad:

KILL (red): CPL > 2x benchmark after 1,000+ impressions → pause immediately HOLD (yellow): CPL between 1x-2x benchmark OR < 1,000 impressions → keep running SCALE (green): CPL < benchmark for 3+ consecutive days → increase budget TEST NEW (blue): All 3 ads killed → return to /ads-brief for new angles

Step 5: Format Report

═══════════════════════════════════════
 AD PERFORMANCE REPORT — {date}
═══════════════════════════════════════

{VERTICAL} Campaign — {days} days active
Budget: ${daily}/day | Total Spend: ${total}
Leads: {count} | CPL: ${cpl} (benchmark: ${bench})

  Ad 1: "{headline}"
    Spend: ${x} | Leads: {n} | CPL: ${x}
    → RECOMMENDATION

  Ad 2: "{headline}"
    Spend: ${x} | Leads: {n} | CPL: ${x}
    → RECOMMENDATION

  Ad 3: "{headline}"
    Spend: ${x} | Leads: {n} | CPL: ${x}
    → RECOMMENDATION

ACTIONS:
  1. {action item}
  2. {action item}

ROI MATH:
  {leads} leads × 20% close rate = {clients} clients
  {clients} × $97/mo = ${mrr}/mo MRR
  Ad spend: ${spend} → Payback: {days} days
═══════════════════════════════════════

Step 6: Save Report & Update Intelligence

Save to: ~/thecalltaker-ops/ads/reports/{vertical}-{date}.json

Update ~/thecalltaker-ops/ads/intelligence.json:

• Update cpls_by_vertical with latest actual CPL
• Update total_spend and total_leads
• Move killed angles to burned_angles
• Move scaled angles to winning_angles

Create directory if needed: mkdir -p ~/thecalltaker-ops/ads/reports

Step 7: Notification

If campaigns are active and have data, send to ntfy SALES topic (tct-sales-63uYsIT9):

AD REPORT: {vertical} | Spend: ${spend} | Leads: {leads} | CPL: ${cpl} | Action: {primary recommendation}

Output to User

Print the full formatted report, then: "Report saved to ~/thecalltaker-ops/ads/reports/{vertical}-{date}.json" "Intelligence updated."