Skip to main content
The Prompt Manager is your central workspace for every prompt GenRank tracks. You define the prompts here, assign them intent categories, toggle web retrieval behavior, and control which prompts are actively being monitored. Every visibility metric, competitor comparison, and trend chart in GenRank flows from the prompts you manage in this view.

​
Adding your first prompt

1

Open Prompt Manager

Navigate to Prompt Research in the left sidebar, then select Prompt Manager. You will see your existing prompt library, or an empty state if this is your first session.
2

Click Add Prompt

Select Add Prompt in the top-right corner of the library view. A prompt creation panel opens on the right side of the screen.
3

Write the prompt

Enter the full prompt text exactly as a user would type it into ChatGPT. Use natural, conversational language — not keyword fragments. For example, write “what is the best accounting software for freelancers” rather than “accounting software freelancers.”
4

Assign an intent category

Choose the category that best describes the intent behind this query. See the intent categories section below for guidance on each option.
5

Configure web retrieval

Decide whether this prompt should force ChatGPT’s web search mode when GenRank runs it. Enable this if you want to test visibility in retrieval-augmented responses. See the web retrieval section below for more detail.
6

Save and activate

Click Save. The prompt is added to your library and begins tracking in the next scheduled run. Newly added prompts are activated by default — you can deactivate any prompt at any time without losing existing data.

​
Intent categories

Every prompt in GenRank belongs to an intent category. Categorizing prompts lets you analyze visibility by funnel stage — so you can see where you dominate and where you are underrepresented.
Discovery prompts reflect early-stage research behavior. Users are exploring a category, not yet evaluating specific options. These are typically “what is the best X,” “top X tools,” or “how do I solve Y” type queries.Examples:
  • “what is the best email marketing platform for small businesses”
  • “top project management tools for startups”
  • “how do companies manage customer support at scale”
Discovery prompts tend to have high volume and are where brand awareness in AI is first established.
Comparison prompts signal that a user is evaluating options. They often name specific products or pairings directly. These are “X versus Y,” “alternatives to X,” or “which is better” queries.Examples:
  • “HubSpot vs Salesforce for a mid-market SaaS company”
  • “best alternatives to Notion for knowledge management”
  • “which is better for e-commerce, Shopify or WooCommerce”
Comparison prompts are critical for competitive intelligence — they reveal whether ChatGPT favors you or a competitor when users are ready to choose.
Transactional prompts reflect high purchase intent. Users are looking for a specific solution for a specific need. These queries often include buyer-context details like company size, budget, use case, or role.Examples:
  • “best CRM for a B2B SaaS startup with a small sales team”
  • “affordable HR software for companies with 50 to 200 employees”
  • “what accounting tool should a freelance designer use”
Transactional prompts typically have the highest conversion relevance and the most direct impact on buying decisions.
Branded prompts include your brand name directly. They capture how ChatGPT describes, positions, and represents your brand when users ask about it specifically.Examples:
  • “what is [Your Brand]”
  • “is [Your Brand] good for enterprise teams”
  • “how does [Your Brand] compare to its competitors”
Tracking branded prompts is essential for understanding your brand entity clarity — how accurately and consistently ChatGPT represents what you do.
Custom is a free-form category for any prompt that doesn’t fit the standard intent types. Use it for industry-specific queries, audience-segment prompts, or experimental tracking that falls outside the standard funnel structure.

​
Web retrieval control

Each prompt has an optional web retrieval toggle. When enabled, GenRank forces ChatGPT to use its web search mode when running that specific prompt. When disabled, the prompt runs without web retrieval, using only ChatGPT’s base model knowledge. This distinction matters because AI responses can differ significantly depending on whether external sources are retrieved. A brand that appears in ChatGPT’s training data may be described differently than one that is primarily known through recent web content. Use cases for enabling web retrieval:
  • Testing whether your recent content, announcements, or product updates are being picked up
  • Understanding how third-party sources like review sites or media coverage shape your brand’s representation
  • Comparing your visibility with and without retrieval to identify where web presence is a factor
Run a paired test: add the same prompt twice, once with web retrieval enabled and once without. Compare the responses in Response Tracking to understand how retrieval changes your brand’s positioning.

​
Bulk import

If you have a large set of prompts — from keyword research, customer interviews, competitor analysis, or export from another tool — you can add them all at once using bulk import.
1

Prepare your prompt list

Format your prompts as a plain text list, one prompt per line. You do not need to include categories or metadata at this stage — those can be assigned after import.
2

Open bulk import

In Prompt Manager, select Import and choose Bulk Add. Paste your prompt list directly into the input field or upload a .txt or .csv file.
3

Review and categorize

GenRank displays a preview of all imported prompts before adding them. Review for duplicates or prompts that need editing, then assign intent categories. You can apply a category to all prompts in the batch or categorize them individually.
4

Confirm import

Click Import to add all reviewed prompts to your library. Prompts that would exceed your plan’s slot limit are flagged — you can upgrade your plan or deactivate existing prompts to free up slots.
Bulk import does not check for exact duplicates across your existing library. Review your current prompts before importing a large batch to avoid tracking the same query multiple times and consuming unnecessary prompt slots.

​
Prompt selection strategy

Start narrow, then expand. Begin with 5–10 high-signal prompts that directly reflect how your category is discovered and compared. Establish a baseline, review your first two weeks of data, and use those results to identify gaps before scaling your prompt set.
When building your prompt library, consider the following principles: Write prompts the way real users ask them. “What’s the best invoicing software for a solo consultant” captures real intent. “Invoicing software consultant best” does not. ChatGPT responds to conversational input, and your prompts should reflect that. Cover the full funnel, not just the bottom. It’s tempting to focus only on high-intent transactional prompts, but discovery and comparison queries often have higher volume and shape brand perception before a purchase decision is made. Include competitor-framing prompts. Queries like “alternatives to [Competitor]” or “[Competitor] vs [Your Brand]” reveal whether ChatGPT positions you favorably when users are actively evaluating options. Revisit your prompt set regularly. Prompt landscapes evolve as products change, new competitors emerge, and user behavior shifts. A prompt library that was accurate six months ago may no longer reflect current conversations in your category. Use Page Scan and Search Console as inputs. Rather than inventing prompts manually, derive them from signals you already own — your existing content and your real search data. See Page Scan and Search Console for details.