AI Analysis

These tools provide intelligent analysis of your advertising data, including summaries, trend detection, and anomaly alerts.

AI Analysis tools are read-only. Editing/updating campaigns/ads and ad-writing features are Coming Soon.

meta_analysis_summary

Generate a comprehensive account performance overview with key metrics and trends.

Description

Provides an executive-level summary of your ad account performance, including headline numbers, comparisons to previous periods, and key highlights. Perfect for daily check-ins or reporting.

Parameters

Parameter	Type	Required	Default	Description
date_preset	string	No	last_7d	Time period for the summary

Date Presets

Preset	Description
today	Today only
yesterday	Yesterday only
this_week	Current week
last_week	Previous week
this_month	Current month
last_month	Previous month
last_7d	Last 7 days
last_14d	Last 14 days
last_30d	Last 30 days

Example Prompts

Give me an account summary

How are my ads performing this week?

Summarize my ad performance for the last 30 days

Response

{
  "success": true,
  "account_name": "My Business Account",
  "period": "last_7d",
  "summary": {
    "spend": 2450.75,
    "impressions": 245000,
    "reach": 98000,
    "clicks": 5635,
    "ctr": 2.30,
    "cpc": 0.43,
    "conversions": 142,
    "cost_per_conversion": 17.26,
    "roas": 3.2
  },
  "highlights": [
    "Spend increased 12% vs previous period",
    "CTR improved from 2.1% to 2.3%",
    "Top performing campaign: Summer Sale with 4.1 ROAS",
    "3 campaigns are currently underperforming their benchmarks"
  ]
}

Response Fields

Field	Type	Description
period	string	The time period analyzed
summary	object	Key performance metrics
highlights	array	AI-generated insights and observations

Use the summary tool for daily standups or weekly reports. The highlights automatically surface the most important changes.

---

meta_analysis_top_movers

Identify campaigns with significant performance changes - gainers and losers.

Description

Compares performance between two periods to identify campaigns that have improved or declined the most. Great for understanding what’s changing in your account.

Parameters

Parameter	Type	Required	Default	Description
metric	string	No	spend	Metric to analyze
comparison_period	string	No	week	Compare vs: day, week
limit	number	No	5	Number of movers per category

Available Metrics

Metric	Description
spend	Total spend
impressions	Total impressions
clicks	Total clicks
cpc	Cost per click
cpm	Cost per 1,000 impressions
ctr	Click-through rate
reach	Unique users reached
frequency	Average times seen per user
conversions	Total conversions
cost_per_conversion	Cost per conversion
purchase_roas	Return on ad spend

Example Prompts

Which campaigns improved the most this week?

Show me top movers by ROAS

What campaigns had the biggest spend changes?

Which campaigns are declining?

Response

{
  "success": true,
  "metric": "spend",
  "comparison_period": "week",
  "current_period": "Jan 8 - Jan 14",
  "previous_period": "Jan 1 - Jan 7",
  "top_gainers": [
    {
      "campaign_id": "123456789",
      "campaign_name": "Summer Sale",
      "current_value": 850.00,
      "previous_value": 520.00,
      "change_percent": 63.5,
      "change_absolute": 330.00
    }
  ],
  "top_losers": [
    {
      "campaign_id": "987654321",
      "campaign_name": "Brand Awareness",
      "current_value": 200.00,
      "previous_value": 450.00,
      "change_percent": -55.6,
      "change_absolute": -250.00
    }
  ]
}

Response Fields

Field	Type	Description
metric	string	The metric analyzed
current_period	string	The current time range
previous_period	string	The comparison time range
top_gainers	array	Campaigns that improved most
top_losers	array	Campaigns that declined most
change_percent	number	Percentage change
change_absolute	number	Absolute change in value

---

meta_analysis_anomalies

Detect unusual patterns, spikes, or drops in performance using statistical analysis.

Description

Uses statistical methods to identify metrics that are significantly different from their historical norms. Helps you catch issues early, like sudden spend spikes or performance drops.

Parameters

Parameter	Type	Required	Default	Description
lookback_days	number	No	14	Days to analyze (max: 90)
threshold	number	No	2.0	Sensitivity in std deviations (lower = more sensitive)
metrics	array	No	default set	Specific metrics to analyze

Default Metrics Analyzed

• spend
• impressions
• clicks
• cpc
• ctr

Example Prompts

Are there any anomalies in my ad performance?

Check for unusual patterns in my campaigns

Is anything weird happening with my ads?

Detect any issues in the last 30 days

Response

{
  "success": true,
  "lookback_days": 14,
  "threshold": 2.0,
  "metrics_analyzed": ["spend", "impressions", "clicks", "cpc", "ctr"],
  "summary": {
    "total_anomalies": 2,
    "critical": 1,
    "warning": 1,
    "healthy_metrics": 3
  },
  "anomalies": [
    {
      "metric": "impressions",
      "date": "2024-01-12",
      "value": 45000,
      "expected_value": 28000,
      "deviation": 2.8,
      "direction": "spike",
      "severity": "critical",
      "is_concerning": true,
      "message": "Impressions were 60% higher than normal. Check if budget was increased or targeting expanded."
    },
    {
      "metric": "ctr",
      "date": "2024-01-13",
      "value": 1.2,
      "expected_value": 2.1,
      "deviation": 2.1,
      "direction": "drop",
      "severity": "warning",
      "is_concerning": true,
      "message": "CTR dropped 43% below normal. Review recent creative changes or audience fatigue."
    }
  ]
}

Response Fields

Field	Type	Description
lookback_days	number	Days analyzed
threshold	number	Sensitivity used
summary	object	Overview of findings
anomalies	array	List of detected anomalies
anomalies[].deviation	number	How many std deviations from mean
anomalies[].direction	string	”spike” or “drop”
anomalies[].severity	string	”critical” or “warning”
anomalies[].is_concerning	boolean	Whether this needs attention
anomalies[].message	string	Human-readable explanation

Interpreting Severity

Severity	Deviation	Meaning
Critical	> 2.5σ	Highly unusual, needs immediate attention
Warning	2.0-2.5σ	Notable deviation, worth investigating

Anomaly detection works best with at least 14 days of data. New accounts or campaigns may not have enough history for accurate detection.

---

Use Cases

Daily Health Check

Start each day with anomaly detection to catch issues early

Are there any problems with my ads today?

Weekly Review

Use top movers for your weekly performance review

Which campaigns improved most this week?

Executive Reporting

Generate summaries for stakeholder updates

Give me an executive summary for last month

Budget Monitoring

Watch for unexpected spend changes

Check for any spend anomalies in the last week

Common Workflows

See how to combine these tools for powerful workflows