Validating Your Tracking with Volume Data

Key Concepts

Phrase Match vs Exact Match: Phrase match counts any prompt containing your keywords in any order. Exact match requires keywords to appear in the specific order you entered. Use phrase match when wording varies; use exact match when word order matters to meaning.

Head Terms: Short, topic-level keywords (one to two words) rather than full prompts. Prompt volumes works best with head terms because longer phrases rarely match how people naturally phrase questions.

Quick Reference

What to Enter in Prompt Volumes

• Use head terms, not raw prompts
• Stick to one or two words when possible
• Three to four word phrases often return little or no data
• If you see no volume, go broader until you get signal

Bulk Analysis (Validation)

Use this to prioritize and compare multiple topics at once.

1. Click Bulk Analysis tab
2. Paste your list of head terms
3. Select Phrase Match or Exact Match
4. Click Analyze

You can also import a .txt file for larger keyword lists.

Results show topics ranked by volume. Filter by platform, region, age, income, or time range.

Single Keyword Analysis (Exploration)

Use this to understand what the market means by a topic.

1. Enter one keyword (do not check bulk analysis)
2. Click Analyze

Available data:

• Volume by model, platform, region, age, income
• Relevant prompt examples (privacy-anonymized, representative language)
• Intent breakdowns (generative, informational, commercial, conversational) with sub-intents like consideration, awareness, direct purchase
• Demographics (country, age, income distribution)
• Keyword hierarchy (sub-terms and related terms at adjustable depth levels)

Passive Monitoring in Answer Engine Insights

Prompt volume data also appears in the Prompts table within Answer Engine Insights.

• Volume is based on the exact topic name you created
• Long or deliberately niche topic names will show low volume
• Customize columns to surface prompt volume data

Note: The prompt volume shown is for the latest complete data period across all models.