What Is Prompt Volume?
Short answer
Prompt volume is an estimate of how often people ask AI assistants about a question or topic, the AI-search version of keyword search volume. No AI platform publishes these numbers, so every figure is a vendor's model: useful for comparing topics, unreliable as an exact count.
Prompt Volume explained
Keyword research runs on search volume: how many times a month people type a query into Google. Teams moving into AI visibility wanted the same number for prompts, and vendors built it. Profound's prompt volumes were among the first. The catch: OpenAI, Google, Anthropic and the other AI companies don't publish how often people ask what, so there's no official figure to check any estimate against.
How vendors estimate it
- Consumer panels: licensed conversation data from opt-in panels of real AI users, scaled up with statistical modeling. Profound says its panel data covers hundreds of millions of prompts a month.
- Search-data proxies: modeling AI demand from keyword volumes and clickstream data on related behavior.
- Systematic querying: running prompts through engines. This shows what engines answer, not what people ask, so it isn't demand data.
Each method has blind spots, and figures from different vendors won't match.
How to use it
- Compare topics, not absolute counts. "Topic A draws about three times the prompts of topic B" holds up better than "1,240 prompts a month."
- Expect prompts to be longer and more specific than keywords. People ask full questions with context, like their company size or budget (long-tail keywords).
- Check estimates against real questions from sales calls, support tickets and site chat.
- Use it to choose which prompts to track for AI visibility, then measure your presence in those answers directly.
How to estimate demand without a tool
Start with the questions customers already ask you. Group them by theme, check related keyword volumes as a rough proxy, and note which themes come up most in sales conversations. That gives you a ranked list of topics, which is what most teams need prompt volume for anyway.
In practice
My rule is direction, not precision. When a dashboard shows an exact monthly count, I tell clients nobody actually knows that number. The useful signal is that one theme draws more demand than another, which helps with prioritizing. I also check the data against questions clients hear on sales calls. That's the most honest prompt-volume data most businesses have.
Common misconception
Prompt volume figures look like search volume, so people read them the same way. They're vendor models built on samples, good for ranking topics against each other and unreliable as counts.
Sources
Free email course: Smarter SEO for the Age of AI
Seven short lessons over three weeks: why clicks are falling even when rankings hold, and how to get AI systems to recommend your brand.
