A practical guide to understanding how different AdPredictor sections relate to each other to maximize your results.
The Data Flow: From Campaigns to Actions
1. You create campaigns in Google Ads
2. We sync your data (every 4-24h based on plan)
3. AI analyzes your performance
4. You get insights and alerts
5. You take action from AdPredictor
Pause keywords, adjust bids, add negatives
6. Next sync shows results of your changes
7. Cycle repeats
Real Example: How Features Work Together
Let's say you run an online shoe store and your "Blue Running Shoes" campaign is underperforming.
Step 1: You Notice a Problem
Location: Dashboard
Your ROAS is 1.8x (you want 2.5x+). The AI has already identified the issue.
Step 2: Investigate Details
Location: Campaigns → "Blue Running Shoes"
- Campaign spent €500 this week
- 150 clicks, only 2 conversions (rate: 1.3%)
- AI says: "Below expected performance. Expected 7-8 conversions, got 2."
Step 3: Check Search Terms
Location: Search Terms
People searched: "cheap running shoes", "discount running shoes", "free running shoes".
AI says: "Block 'cheap' and 'free' as negative keywords. Estimated savings: €120/month."
Step 4: Take Action
From AdPredictor, you can:
- Pause keywords that don't convert (all plans)
- Add "cheap" and "free" as negative keywords (Professional+)
- Adjust bids on winning keywords (Professional+)
Step 5: Check Results
After your next sync, your ROAS should improve as wasted spend is eliminated.
Common Workflows
Workflow 1: "I Want Better Quality Scores"
- Go to Keywords → filter by low Quality Score
- Read Insights → AI explains which ads need fixing
- Check Search Terms → find keyword-search mismatch
- Pause low-quality keywords or adjust bids
- Next sync: quality scores should improve
Workflow 2: "I'm Losing Money on Bad Keywords"
- Go to Keywords → sort by cost (highest first)
- Look for keywords with "clicks but 0 conversions"
- Check Search Terms → see what people actually searched
- Go to Insights → AI suggests "Add negative keywords"
- Pause losing keywords and add negatives from AdPredictor
Workflow 3: "I Want Quick Wins"
- Go to Dashboard → "Quick Wins" section
- Each quick win links to the problem area
- Implement each recommendation
- Monitor results via ROAS trending
Workflow 4: "Something's Wrong and I Don't Know What"
- Go to Alerts
- Read the alert (e.g., "Spend spiked 150% today")
- Click the alert → get context
- Decide: Was this intentional? If not, take action
- If critical, use the AI Assistant to ask "Why did my spend spike?"
Which Features Depend on Plan
| Feature | Starter | Professional | Enterprise |
|---|---|---|---|
| View campaigns, ads, keywords | ✓ | ✓ | ✓ |
| Pause/enable keywords | ✓ | ✓ | ✓ |
| Adjust bids & budgets | — | ✓ | ✓ |
| Add negative keywords | — | ✓ | ✓ |
| Bulk actions | — | ✓ | ✓ |
| Automated rules | — | ✓ | ✓ |
| AI Assistant | — | ✓ | ✓ Priority |
| Anomaly detection | — | ✓ | ✓ |
| PDF reports | — | ✓ | ✓ |
| AI insights | 5 per analysis | Unlimited | Unlimited |
| Data history | 30 days | 90 days | 12 months |
Tips for Using Features Together
- Tip 1: Start at Dashboard — see the big picture → dive into specifics
- Tip 2: Trust the AI's prioritization — implement the top 3-5 insights per week
- Tip 3: Check Search Terms weekly — find customer intent and block irrelevant traffic
- Tip 4: Use the AI Assistant for custom questions — don't guess, ask!
- Tip 5: Review your data regularly — weekly: Dashboard + Insights. Monthly: trends vs. actuals.
AdPredictor is a connected system. Every feature feeds into AI analysis, which surfaces insights. Your job is to act on recommendations and measure results.
Common Mistakes & How Features Catch Them
| Mistake | Feature That Catches It | What It Says |
|---|---|---|
| Bidding on irrelevant keywords | Search Terms + Insights | "People searched X, but you're bidding Y" |
| Wasting budget on zero-converting keywords | Insights | "50 clicks, 0 conversions → pause this keyword" |
| Missing opportunities | Insights | "This keyword converts at 20%, but bid is low" |
| Campaign suddenly broken | Alerts | "Spend spiked 150%, conversions dropped 60%" |