Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.topify.ai/llms.txt

Use this file to discover all available pages before exploring further.

Prompts are the foundation of Topify.ai’s performance tracking. Each prompt represents a question that real users ask AI assistants — and that Topify.ai sends to AI providers daily on your behalf.

​
What is a prompt?

A prompt is a natural-language query that your target audience might type into an AI search engine. For example:
  • “What is the best project management tool for startups?”
  • “Which CRM has the best automation features?”
  • “How do I choose a web hosting provider?”
Topify.ai sends each active prompt to all monitored AI providers (see Integrations for the full list) and collects the responses. These responses are then analyzed to determine whether your brand was mentioned, how it was portrayed, and what other brands appeared.

​
Why prompts matter

Prompts are the lens through which you monitor your brand’s AI visibility. Choosing the right prompts means you’re tracking the conversations that matter most to your business. Each prompt generates a set of metrics:
MetricWhat it tells you
VisibilityHow often your brand appears in responses to this prompt (as a percentage across providers)
SentimentHow positively your brand is described when it appears
PositionWhere your brand appears in the response relative to competitors
CVRCitation Visibility Rate — a composite score (0—100) blending visibility, sentiment, and position into a single quick-glance prominence score for this prompt
VolumeHow many AI responses have been collected and analyzed

​
Types of prompts

Prompts that Topify.ai tracks daily. Responses are collected and analyzed each day, and metrics appear on your dashboard.

​
How AI search volume is estimated

Volume estimates how often a question is being asked across AI platforms. Topify extracts keywords from each prompt, queries Google Ads search volume data as a baseline, then applies AI-specific adjustments — including intent classification and topic coherence — to estimate AI-specific query volume.

​
Prompt intent

Every prompt carries an intent that influences how metrics are weighted:
IntentDescriptionExample
CommercialEvaluating products or services before a purchase”Best CRM for small businesses”
TransactionalReady to take action or buy”Sign up for Salesforce free trial”
InformationalSeeking knowledge or understanding”What is a CRM?”
NavigationalLooking for a specific brand or page”HubSpot pricing page”
Commercial and transactional prompts typically carry higher business value because they represent users closer to a purchasing decision.
For a walkthrough on adding and organizing prompts, see Managing prompts.

​
FAQ

​
How should I structure my prompt set?

Cover the four intent types — informational, commercial, transactional, navigational — so you see the full buyer journey, not just one slice. Then layer in supporting prompts that catch the long-tail variations of each core question. For example, alongside “best CRM” track “best CRM for 5-person sales team”, “best CRM under $50/seat”, and “best CRM for B2B SaaS startups.” Real buyers ask the long-tail versions far more than the generic ones, and AI answers can vary a lot between them.

​
Can I track prompts in multiple languages?

Yes. Each project has a default language, but you can add prompts in any language from Tracking > Prompts > + Add Prompts. AI providers respond in the prompt’s language and metrics are calculated the same way. If you operate in several markets, consider one project per language so the dashboards stay clean.

​
How many prompts should I track?

Start with 15–30 active prompts that cover your core products, top use cases, and common buyer questions. Most projects don’t need more than that to get a clear picture. Refine over time — drop the noisy ones, add new ones as themes emerge from your data.