Prompt types and what they reveal
Use the right mix for the right insight.
Last updated: 2026-05-05
Prompt types and what they reveal
Pranas classifies every prompt into one of five types. Each type measures a different stage of the buyer journey, and segmenting your dashboard by type is the single fastest way to turn raw visibility numbers into actionable insight. A brand can rank #1 on Comparison prompts but be invisible on Unaided — that's a very different problem from being invisible across the board.
Prompt type reference
| Type | Example | Funnel stage | What it measures | Target share of library |
|---|---|---|---|---|
| Unaided | "What's the best CRM for small teams?" | Top | Organic discovery and true recall | 40% |
| Aided | "Tell me about Pranas." | Any | Whether the AI knows you accurately | 15% |
| Comparison | "Pranas vs Writesonic — which is better?" | Bottom | Head-to-head consideration | 20% |
| Problem-Solving | "How do I track brand visibility in ChatGPT?" | Middle | Category-language mapping | 15% |
| Feature / Use-Case | "Which tool monitors AI Overview citations?" | Middle | Capability-level positioning | 10% |
The five prompt types
1. Unaided
The user does not mention your brand or any competitor. Example: "What's the best CRM for small teams?"
This is the gold standard. If an AI model spontaneously mentions you here, you have organic visibility — the same kind of brand recall that traditional advertising chases. Unaided is what most dashboards (Top Brands, SOV, AI Score) default to because it's the cleanest signal of true market presence.
2. Aided
The user mentions your brand by name. Example: "Tell me about Pranas."
Aided prompts confirm the AI knows you exist and can describe you accurately. Low Aided scores often mean the model has stale or missing training data about your brand — a content/SEO problem. High Aided + low Unaided means the AI knows you but doesn't recommend you.
3. Comparison
The user pits two or more brands against each other. Example: "Pranas vs Writesonic — which is better?"
Comparison prompts are bottom-of-funnel. Winning these means you appear in head-to-head consideration. Losing them with a specific competitor highlights exactly which positioning narrative you need to fix.
4. Problem-Solving
The user describes a pain point without naming any product. Example: "How do I track my brand's visibility in ChatGPT?"
This type measures whether AI models map your category language onto your brand. It's where category-defining content (pillar pages, "what is" articles) pays off.
5. Feature / Use-Case
The user asks about a specific capability. Example: "Which tool monitors AI Overview citations?"
Strongest signal for product-marketing teams. If you have a unique feature but don't appear here, your messaging hasn't reached the AI's training data yet.
How to use the segmentation
Every dashboard widget that displays SOV, Top Brands, or Visibility supports a prompt-type filter. The default view is Unaided because it's the least biased. Switch to:
- Comparison when prepping for a competitive launch
- Problem-Solving when planning content marketing
- Feature/Use-Case when launching a new feature
- Aided to QA whether the AI has up-to-date facts about you
Recommended balance
For a healthy library, aim for roughly:
- 40% Unaided
- 15% Aided
- 20% Comparison
- 15% Problem-Solving
- 10% Feature/Use-Case
The auto-generator targets this mix by default. You can rebalance manually on the Prompt Management page.
Related articles
- Generate prompts automatically
- Search volume tiers
- Share of Voice (SOV)
- Top Brands widget