Smart Alerts Introduction

Smart Alerts are AI-generated notifications that identify issues, opportunities, and anomalies in your Google Ads campaigns. Powered by Claude Sonnet 4, each alert includes a severity rating (1-10), detailed explanation, and recommended action - helping you fix problems before they cost money.

What You'll Learn

Time needed: 8 minutes to read

What Are Smart Alerts?

Definition: Automated notifications generated by AI when it detects performance issues, budget problems, technical errors, or optimization opportunities in your Google Ads campaigns.

Example:

Severity: 8/10 - High Priority
Category: Budget Issues
Title: Daily budget exhausted before 2 PM

Your campaign "Summer Sale 2025" spent its entire $100 daily
budget by 1:45 PM EST for the past 3 days. You're missing
evening traffic (40% of your conversions happen after 6 PM).

Recommended Action: Increase daily budget to $150 or adjust
ad schedule to focus spend on 6-11 PM when conversion rate
is highest (5.2% vs 2.1% daytime average).

[Screenshot: Example alert card showing high severity budget alert] AI alerts highlight problems and suggest fixes

How Smart Alerts Work

The AI Analysis Process

Step 1: Data Collection

Step 2: AI Review

Step 3: Alert Generation

Step 4: Presentation

[Screenshot: Alert flow diagram showing data → AI → alerts → action] From data to action in 4 steps

When Alerts Are Generated

Automatic Analysis

Triggers:

Timing:

2:00 AM: Auto-sync runs
2:15 AM: Sync completes
2:15 AM: AI analysis begins
2:18 AM: Alerts generated
9:00 AM: You check client and see new alerts

Frequency: Daily for active clients with auto-sync enabled

Manual Force Analysis

When to use:

How to force:

  1. Go to Client Details → AI Alerts tab
  2. Click "Force Full Analysis" button
  3. Wait 2-3 minutes
  4. New alerts appear

Cost: Uses Claude API tokens (~1,000 tokens per analysis)

[Screenshot: Force Full Analysis button location] Manual trigger for immediate AI review

Understanding Alert Severity

The 1-10 Scale

How AI assigns severity:

Severity Levels:

9-10: Critical (Red) 🔴

7-8: High (Orange) 🟠

5-6: Medium (Yellow) 🟡

3-4: Low (Light Yellow) 🟡

1-2: Informational (Blue) 🔵

[Screenshot: Alert severity color coding] Color-coded alerts for quick prioritization

Alert Categories Explained

Budget Issues 💰

What it catches:

Example:

"Campaign spending only $20 of $100 daily budget. Increasing bids by 30% could capture more traffic without exceeding budget."

Why it matters: Budget problems waste money or miss opportunities

Performance Problems 📉

What it catches:

Example:

"Average CPC increased from $3.50 to $5.20 (49% jump) in last 14 days with no improvement in conversion rate."

Why it matters: Performance declines directly impact ROI

Technical Issues ⚠️

What it catches:

Example:

"3 ads disapproved for policy violation. Campaign has only 1 active ad remaining, limiting impression share."

Why it matters: Technical issues can stop campaigns entirely

Strategy Recommendations 🎯

What it catches:

Example:

"85% of conversions happen Mon-Fri 9 AM-5 PM. Consider ad scheduling to increase bids during these hours."

Why it matters: Strategic changes improve efficiency

Quality Score Problems ⭐

What it catches:

Example:

"12 keywords have Quality Score of 3 or lower, causing 2x higher CPCs. Improving ad relevance could save $800/month."

Why it matters: Quality Score directly affects CPC and ad position

Conversion Tracking 🎯

What it catches:

Example:

"No conversions recorded in last 7 days despite 450 clicks. Check if conversion tracking pixel is installed correctly."

Why it matters: Can't optimize what you can't measure

Geographic Performance 🌍

What it catches:

Example:

"California generates 65% of conversions at 40% lower CPA than other states. Consider increasing California budget."

Why it matters: Location targeting impacts efficiency

Bidding Strategy 💡

What it catches:

Example:

"Manual CPC campaign has high conversion volume. Switching to Target CPA could automate bid optimization and reduce workload."

Why it matters: Right bidding strategy improves results and saves time

Why Smart Alerts Beat Manual Monitoring

The Manual Way (Without Alerts)

Your workflow:

  1. Log into Google Ads
  2. Open each campaign
  3. Check metrics manually
  4. Look for anomalies
  5. Compare to previous periods
  6. Spot patterns (maybe)
  7. Decide what to fix
  8. Repeat for 20 campaigns

Time required: 30-45 minutes per client Frequency: Weekly if you're diligent Issues caught: Only obvious problems

Problems:

The Smart Alerts Way

What happens:

  1. AI analyzes all campaigns automatically
  2. Spots patterns humans miss
  3. Compares to 30-day baseline
  4. Identifies issues before they escalate
  5. Provides specific recommendations
  6. Prioritizes by severity
  7. Updates daily

Time required: 2 minutes to review alerts Frequency: Daily, automatically Issues caught: Everything significant

Benefits:

[Screenshot: Side-by-side comparison of manual vs AI monitoring] AI finds problems faster and earlier

How Alerts Appear

In Client Details

Location: Client Details → AI Alerts Tab

What you see:

Interactions:

[Screenshot: AI Alerts tab showing multiple alerts] All client alerts in one organized view

On Client List

Location: Client List → Alerts Column

What you see:

Quick scan:

Client Name          Alerts
ACME Corp           🔴 3    (3 critical/high alerts)
Bakery Inc          🟡 5    (5 medium alerts)
Law Firm LLC        🟢 0    (no alerts)

Benefit: Spot problem clients instantly

[Screenshot: Client list showing alert badges] Color-coded badges for quick scanning

In Global Alerts View

Location: Main Navigation → Alerts

Shows:

Use case: Agency-wide alert monitoring

[Screenshot: Global alerts view] See all client alerts in one place

Alert Lifecycle

Creation

Trigger: AI analysis completes after sync

Content:

Active State

Visible:

Actions available:

Resolution

Happens when you:

Result:

Alert History

What it tracks:

Where to find: AI Alerts tab → Alert History

Use cases:

[Screenshot: Alert history view] Full record of AI recommendations and outcomes

Force Full Analysis

When to Use It

Scenarios:

How It Works

Process:

  1. Click "Force Full Analysis" button
  2. System triggers immediate AI review
  3. Uses latest synced data (doesn't sync new data)
  4. Generates new alerts in 2-3 minutes
  5. Replaces old alerts with fresh analysis

Cost:

What Gets Analyzed

Data reviewed:

Output:

[Screenshot: Force Full Analysis in progress] Manual trigger for immediate AI insights

AI Learning from Your Feedback

How Feedback Improves Alerts

When you accept an alert:

When you decline an alert:

Long-term benefits:

Privacy of Feedback

Important:

Tips & Best Practices

Daily Alert Review

Morning routine:

  1. Check global alerts view (2 minutes)
  2. Sort by severity (critical first)
  3. Open client details for critical alerts
  4. Take action on high-priority items
  5. Schedule medium-priority for later

Time investment: 5-10 minutes Result: Catch problems within 24 hours

Use Severity as a Guide

Don't ignore low-severity alerts - they're future problems Do prioritize high-severity - these cost money now

Strategy:

Accept vs Decline Thoughtfully

Don't blindly accept all alerts - AI needs honest feedback

Good reasons to accept:

Good reasons to decline:

Your feedback teaches the AI what matters to you

Combine with Other Data

Alerts + Performance Charts:

Alerts + Sync Logs:

Alerts + AI Chat:

[Screenshot: Using alerts with charts] Alerts point to problems, charts show details

Force Analysis Strategically

Don't overuse (costs tokens, doesn't add value if data is fresh)

Good times to force:

Bad times to force:

Common Questions

Q: How often are alerts generated? A: Automatically after each daily sync. You can also force analysis manually anytime.

Q: Do alerts replace manual Google Ads checking? A: They catch 95% of issues automatically, but you should still review Google Ads directly for strategy changes and new campaign setup.

Q: What if I disagree with an alert's severity? A: Decline it and provide feedback. The AI learns from this and will adjust future severity ratings.

Q: Can I disable alerts for a specific client? A: Not currently, but you can filter them out. Alert customization is planned for future release.

Q: How many alerts are too many? A: Healthy accounts have 3-7 medium/low alerts. 10+ high-severity alerts suggests major campaign issues need attention.

Q: Do alerts consider my business goals? A: Yes, if you've filled out the business profile. AI uses your target CPA, ROAS goals, and budget to assess severity.

Q: What happens to old alerts? A: They move to Alert History after 7 days or when issue is resolved. History is kept for 90 days.

Q: Can I export alerts? A: Not currently. Export functionality is planned for future release.

Troubleshooting

Problem: No alerts appearing

Diagnostics:

  1. Check if client has been synced recently
  2. Verify sync completed successfully
  3. Check if Force Analysis is available

Common causes:

Solution:

Problem: Too many low-priority alerts

Diagnostics:

  1. Check client campaign health in Google Ads
  2. Review Quality Scores
  3. Check for disapproved ads

Cause: Campaign has many small issues

Solution:

Problem: Alert says issue is fixed but still showing

Cause: Alert hasn't been resolved yet

Solution:

  1. Accept the alert to mark it resolved
  2. Or wait for next automatic analysis
  3. Or force full analysis now

Alerts don't auto-dismiss - you must accept/decline to clear them

Next Steps

Master AI-powered optimization with these guides:

Ready to use Smart Alerts? Go to any Client → AI Alerts tab

Want deeper analysis? Force Full Analysis for immediate insights

Need help interpreting an alert? Ask AI Chat for clarification

Last Updated: October 4, 2025