What is prompt volume? Definition and how it is estimated

Prompt volume is the estimated number of times people ask AI engines a prompt or topic. Why no engine publishes it, how tools estimate it and how to use it.

  • Alvaro Peña de Luna Alvaro Peña de Luna
  • date icon

    Wednesday, Sep 30, 2026

Prompt volume is the estimated number of times people ask AI assistants or AI search engines a given prompt, or a group of prompts on the same topic, over a period of time, usually a month. Because no AI engine publishes these figures, prompt volume is always an estimate, never a measured count.

It is the AI search counterpart of keyword search volume, and it answers the same planning question: which questions are worth tracking and optimizing for first?

Why prompt volume is always an estimate

Search engines report keyword volumes from their own logs. AI engines do not share anything similar: ChatGPT and Perplexity do not publish how often each question is asked. Google includes AI Overviews and AI Mode traffic in the overall web search data of Search Console, as its guide to AI features and your website explains, but it does not report prompt counts for AI answers.

Prompts are also harder to count than keywords, even with access to the data:

  • They are long and unique. People write full sentences with their own context, so the exact same prompt rarely repeats.
  • They are conversational. A question often arrives as the third or fourth turn of a conversation, and only makes sense with what came before.
  • They fan out. One prompt can trigger several searches behind the scenes (see query fan-out), so the prompt and the searches it causes are different things.

How tools estimate prompt volume

Tools that show prompt volume usually rely on one of two approaches, or a mix of both:

  • Panel data. A sample of users who agreed to share their activity (for example through a browser extension or an app) is observed, and the counts are extrapolated to the whole user base.
  • Modelling from search data. Prompts are mapped to related keywords, and keyword volumes are converted into an estimated prompt demand for the topic.

A panel extrapolation follows this logic:

Estimated prompt volume = prompts observed in the panel × (total users of the engine ÷ users in the panel)

Hypothetical example. A panel of 100,000 users asks questions about a topic 500 times in a month. If the engine had 10,000,000 users with the same behaviour, the estimate would be 500 × (10,000,000 ÷ 100,000) = 50,000 prompts a month. Every part of that calculation carries error: the panel may not represent the engine’s users, the total user count is itself an estimate, and grouping prompts into a topic is a judgment call.

How to use prompt volume

  • Compare topics, not single prompts. Estimates are more stable for a cluster of related questions than for one exact wording.
  • Use it to rank, not to forecast. “Topic A seems larger than topic B” is a reasonable reading. “We will get 50,000 impressions” is not.
  • Cross-check with your own data. Keyword volumes, your Search Console queries, and the questions customers ask sales and support are real demand signals you already have.
  • Weigh intent over size. A small topic where buyers compare vendors can be worth more than a large informational one.

Common mistakes

  • Presenting an estimate as a measured count in a report.
  • Comparing prompt volumes from two tools that estimate in different ways.
  • Dropping a topic because its estimated volume is low, when it is close to a purchase.
  • Treating keyword search volume and prompt volume as the same number.

Prompt volume and Mencoro

Prompt volume is not one of the metrics in Mencoro’s methodology. That methodology is explicit that real conversations with AI assistants are private, so it shows how engines answer the questions you choose, not what every person asks or sees. You decide which prompts matter (you can write them or generate suggestions), and Mencoro runs them on ChatGPT, Perplexity, Google AI Overview and Google AI Mode to measure Coverage, Favorability, Share of voice and positions. Read How Mencoro works for the details, and see prompt tracking for how to build a prompt set. For a quick look at how one prompt is answered today, try the free AI visibility checker.

FAQ

Frequently asked questions

No. AI engines do not publish how many times each prompt is asked. Google Search Console includes traffic from AI Overviews and AI Mode in its overall web search data, but it does not report prompt counts for AI answers. Any prompt volume you see in a tool is an estimate.
They are directional at best. They come from panels of users or from models built on keyword data, and both carry sampling and modelling error. Use them to compare topics with each other, not as exact counts.
No. Keyword search volume counts searches for a short query in a search engine, based on the engine’s own data. Prompts are longer, more varied and often part of a conversation, and no engine reports them, so prompt volume has to be estimated and is usually grouped by topic.
No. You can build a strong prompt set from your customers’ language, your sales and support questions and the searches that already bring you traffic. Prompt volume estimates can help you rank topics, but they are not a prerequisite.

Start tracking your brand in AI search today

Monitor how AI engines cite your brand, track keyword positions, and benchmark against competitors, all in one platform.