Available on: all plansFind it in: Analytics > Prompt Volumes
The problem with traditional keyword research
When someone wants to find the best corporate credit card option, they used to Google “best corporate credit card.” Now they ask their preferred AI chat: “What’s the best corporate credit card for small teams?” Traditional keyword research tools capture what people type into search bars, not how they hold conversations with AI. The language is different, the intent is different, and the questions are longer and more specific. Without visibility into real AI prompting behavior, teams face three compounding problems:- Misdirected tracking: monitoring prompts that reflect SEO assumptions, not actual user questions.
- Misaligned content: writing for search engines while users are talking to AI assistants.
- Missed opportunities: finding out about rising topics only after competitors have already moved on them.
Profound’s solution
Prompt Volumes is a proprietary dataset built by aggregating, cleaning, and modeling real conversations between users and AI answer engines. It transforms raw AI conversation data into structured, searchable insights about what your market is actually asking, so you can:- Validate whether the topics you’re tracking reflect real user demand
- Discover the exact phrasing people use when asking about your category
- Understand user intent across informational, commercial, generative, and conversational queries
- Identify which regions are driving specific queries
- Prioritize content and tracking strategy on signal, not assumption
How Prompt Volumes work
Prompt Volumes is a dataset built by licensing AI answer engine conversations, cleaning the data, applying probabilistic modeling, and visualizing the results as intuitive charts and tables your team can act on.All data is anonymized, aggregated, scrubbed of PII, and fully compliant with GDPR and CCPA.
Key features
Keyword search and Bulk analysis
Enter a short, broad keyword to query the dataset and see how often that topic appears across real AI conversations. You can also compare multiple keywords at once using Bulk analysis (available on the Enterprise plan), which ranks topics by volume and helps you prioritize what to track before you invest in it. Keywords can be searched using phrase match (any word order) or exact match (specific word order). Start broad and use the results to narrow down into more specific language.Volume and trend visualization
The Overview tab shows how prompt volume for your keyword changes over time, alongside a breakdown of conversations by intent and a list of related terms worth exploring. Use it to spot rising topics and understand the specific language your audience is using.Intent analysis
The Conversation Share by Intent chart on the Overview tab breaks down why people are asking about a topic, classifying prompts into informational, commercial, generative, and conversational categories. This helps you match content to where users are in their journey.Regions
The Regions tab shows where prompts about your keywords come from, as a heat map and a per-region volume table. Note that coverage is strongest in the US, UK, and select European markets.Keyword lists
Organize related keywords into named lists to track topic clusters over time and keep your research organized by project or audience segment.Relevant prompts sample
Discover a sample of real user prompts that caused an AI answer engine to cite a specific page on your (or a competitor’s) domain in its response. Compare your citations with competitors and explore which exact URLs on your site are being cited and for what kinds of questions.Prompt Research Reports
Prompt Research Reports provide data-driven gap analysis, comparing what real users are asking AI engines against what you’re currently tracking in Answer Engine Insights. Use them to get a data-backed starting point, surface what you may be missing, or understand where your customers are in their purchasing journey.Historical data
Prompt Volumes historical data availability depends on your region and platform. For the US, ChatGPT, Gemini, and Perplexity data goes back to January 2025. For all other regions Profound currently tracks, data is available from January 2026. This means you can query any keyword (even one you just added) and view historical volume trends from January 2025 onward for the US, or January 2026 onward for other regions.Data coverage varies by region. For example, in the US, data is currently available for ChatGPT, Gemini, and Perplexity, while Norway only has ChatGPT.