> ## Documentation Index
> Fetch the complete documentation index at: https://docs.heralded.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Your market

> How a Snapshot measures its category's Google searches, how many open with Google's AI summary, the questions buyers ask LLMs, and their ad value.

A Snapshot's **Your market** section shows how often people search Google for what your category sells, how many of those searches Google answers with an AI summary, how many questions buyers ask LLMs, what all of it is worth in ads, and who the AI answers name. Google's figures are measured. The LLM questions and the ad value are estimates.

## Where it is measured

The figures cover the market you chose when you ran the Snapshot.

* A country is measured in that country.
* **Global** is measured in the United States.
* **Europe** is measured in 10 countries: Germany, the United Kingdom, France, Italy, Spain, the Netherlands, Poland, Sweden, Denmark and Romania. Europe as a whole searches more, so its figures are floors. A country whose data can't be read drops out, and the section says how many markets answered.

## Category demand

Category demand is how many times a month people search Google for what your category sells.

1. Heralded writes search terms for your category, and for the wider kind people start from where your profile names one, in each country's languages and in English.
2. Google Ads Keyword Planner returns the related searches people make in that country, in any language, with each one's monthly volume and price per click. Each country is asked once.
3. An AI relevance check reads the most searched of them against your category, and keeps those that match.
4. Searches that name your brand or a competitor are left out, and so are generic searches Keyword Planner counts under a brand's search.
5. Keyword Planner gives close variants, such as a plural and its singular, one shared figure. Each such group counts once.

Below 50 searches a month, demand reads as low. When no country's Keyword Planner answers, demand is unknown, and the section shows who the AI answers name without volumes.

Heralded researches the category before it plans your topics, so the same searches shape them. A Snapshot of the same site, market and category in the same calendar month reuses the first one's research, so it shows the same figures. A brand you start tracking from that Snapshot in the same month reuses it too.

## Ad value

Ad value is what Google Ads would charge for the clicks the top result gets across your category's demand: its Google searches and the questions buyers ask LLMs, times 0.226 clicks each, times the average price per click. It shows what this demand is worth to the businesses in it. It is not revenue, and not the traffic you would get.

A search is a results page someone saw, not a click. The 0.226 is First Page Sage's click-through rate for the first result on a page without an AI answer, for US searches from March to August 2026, desktop and mobile together. An LLM answer has no first result, so its question counts as a search would.

The price per click is the average over the searches Google prices, weighted by their volume.

Under the figure, the section shows how much of that demand AI answers: the searches that open with Google's AI summary and the questions buyers ask LLMs, as a share of all of it.

Google prices clicks in US dollars. Heralded converts them to euros at the European Central Bank's last published reference rate and keeps that rate with your report. The ECB publishes no rate on weekends and TARGET holidays, so the last working day's rate applies then.

## What you capture

Under each figure, the section estimates how much of it you capture today, and what share that is of what the top spot would get.

* **Google searches.** For each of the 20 top searches, Heralded reads where your own site ranks among the first 10 results. A page without an AI summary gives you the click-through rate of that position. A page with one gives you the top spot's 0.226 when its AI summary names you or links to your site, shared evenly with the competitors the same summary names, as a share of voice is. Otherwise it gives you the lower rate First Page Sage measures at your position on pages with an AI answer. The rates are weighted by each search's volume and applied to all the category's searches.
* **LLM questions.** The LLM questions whose answers recommend you, at your share in the **LLMs** panel.
* **Ad value.** Those Google clicks plus those LLM questions, priced like the ad value.

Each share is of what the top spot would get, so none can pass 100%. These figures count the top spot's equivalent you hold, not visits to your site. Where Heralded has no rankings for the searches, the section shows only the LLM figure. Reports issued before Heralded estimated what you capture show none.

| Position | Without an AI answer | With an AI answer |
| - | - | - |
| 1 | 0.226 | 0.036 |
| 2 | 0.102 | 0.014 |
| 3 | 0.058 | 0.008 |
| 4 | 0.037 | 0.005 |
| 5 | 0.025 | 0.003 |
| 6 | 0.018 | 0.003 |
| 7 | 0.013 | 0.002 |
| 8 | 0.009 | 0.002 |
| 9 | 0.007 | 0.001 |
| 10 | 0.005 | 0.001 |

The rates are First Page Sage's for US searches from March to August 2026, desktop and mobile together.

## AI summary share

Heralded opens Google's results for the 20 most searched of these searches, each in the country it is searched in, and checks whether the page opens with Google's AI summary, the AI Overview. Product cards inside an AI summary count with it. A shopping carousel or a shopping ad on its own is a list of products, not an AI answer, so it doesn't count. The share counts each search by its volume, so a search made more often weighs more. It is a share of top searches, over the pages Heralded could read.

The section shows the five most searched meanings among them. Searches that mean the same thing, such as one search in two spellings or in two languages, share one row with their volumes added up, the way keywords group in the Command Center. Each row shows where your site ranks, whether an AI summary opened on any of its searches, whether it names you, and which competitors it names.

## LLM questions

Heralded has no count of the questions buyers ask LLMs about your category, so it estimates them: category demand times one ratio for each LLM a Snapshot reads, added up.

| LLM | Questions for each Google search | Source |
| - | - | - |
| ChatGPT | 0.15 | The midpoint of Ahrefs Brand Radar's AI-adjusted 0.115 and OpenAI's statement of 2 November 2026 that ChatGPT's search-like volume is "roughly one-fifth" of Google's |
| Google AI Mode | 0.10 | Ahrefs Brand Radar, AI-adjusted volume |
| Gemini | 0.011 | Ahrefs Brand Radar, AI-adjusted volume |

Claude has no published ratio, so it adds none. The same ratios apply in every country and industry. Heralded set them on 9 November 2026. Your report keeps the version of the ratios it used, so a later change never moves a report already issued.

## Who the answers name

The **Google** panel shows the share of your Snapshot's AI Overview answers to its buyer questions that name each brand. The **LLMs** panel shows the share of its ChatGPT, Gemini and Google AI Mode answers that recommend each brand, each LLM weighed by its share of the LLM questions. Each panel shows you and the three competitors on your report with the highest shares, each with its [95% range](https://docs.heralded.ai/methodology/ranges.md). A brand the answers name that isn't one of your report's competitors takes no row.

A Snapshot reads up to 40 answers per engine, so its ranges are wide. The section calls the race close while your range overlaps a competitor's, and says you lead or trail only once the ranges separate.

When a competitor is named in Google's AI summaries more often than you, the section estimates what matching it would reach: how many more searches a month would show you in the AI summary if you were named as often as the most-named competitor. It is your category's searches that open with an AI summary, times the gap between its share and yours. It counts appearances in the AI summary, not clicks or visits.

## Weekly full Snapshots

Your weekly full Snapshot uses your brand's original market and **All buyers** audience. Category demand comes from the existing research for that market and audience, repriced monthly. The AI summary share and the top five meanings come from the report's tracked keyword checks, labelled **tracked searches**, grouped by the English wording your keywords carry in the Command Center. The panels use that market and audience's answers from the week.

Searches naming your brand don't count toward the AI summary share or appear in the top searches.

The section buys no additional keyword research and changes neither your Heralded Score nor your topics' demand weights. If research is not available yet, the section shows the tracked searches and panels without category volumes.

Your first report can arrive before its keyword checks are ready. In that case, **Your market** is unavailable for that report. The next weekly full Snapshot uses its own checks.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.