# The free AI visibility report

> Run up to three of your buyers' questions across the AI engines for free and get a scorecard you can share. It shows how often answers name you, who they name instead and which pages they cite.

sourcesignal documentation, Free tools. Updated 2026-10-18. The page for people: https://www.sourcesignal.ai/docs/free-tools/the-free-ai-visibility-report. Every article: https://www.sourcesignal.ai/llms.txt

The free AI visibility report (the scorecard) shows how often AI engines name your brand when buyers ask the questions you choose. You need no account and no card. It is a good first look before you sign up.

You give it up to three questions. It runs each one many times on each engine and publishes the result on a page with its own link. Every rate on that page is taken over those runs, and every number shows its own sample size.

## What you need

- **Up to three questions.** Write what a potential customer might ask, not your brand name.
- **Your brand name and domain.** These identify your brand in answers and citations.
- **A market.** The country the questions are asked from: Worldwide, United States, United Kingdom, Ireland, Australia, Canada or New Zealand.
- **Competitors (optional).** "Find competitors automatically" is on by default. It picks up to 3 of your closest competitors from your website and checks for them in the same answers. Switch it off to name up to 3 yourself.
- **A work email.** The link is sent there.

## How to run your scorecard

1. Open the [free report form](https://www.sourcesignal.ai/free-report). The "Get free report" button in the site header leads there too.
2. Type your first question. Press "+ Add another question" for a second and a third.
3. Enter your brand name and domain. The market follows your domain's country code until you pick one yourself, and your pick is kept.
4. Leave "Find competitors automatically" on, or switch it off and type competitor names. Press Enter or a comma to add one.
5. Enter your work email and press "Run my free scorecard".
6. You land on your scorecard's own page, which shows the run as it happens.

## While it runs

The page reads "Scorecard in progress" and shows a progress board: an elapsed clock, how many checks are complete, and one row per engine with its phase: "Waiting for a slot", "In the engine queue", "Sampling now", then "Wrapping up" once its samples are all in, and "Done" or "Unavailable (provider)" at the end.

- The page updates itself every few seconds and turns into the finished scorecard when the run ends.
- The page tells you when it is ready, and an email follows.
- A limited number run each day. When the day's allowance is spent, the page reads "Your request is in the queue" and the run starts when the next day's allowance opens.
- When it is ready, an email arrives with a subject that starts "Your AI visibility scorecard is ready" and a "View your scorecard" button.

If a run does not complete, the page reads "Scorecard failed". Nothing is published and no numbers are made up. You can start a trial or contact support from there.

## The engines it runs

Today a scorecard runs ChatGPT, Gemini, AI Overviews, AI Mode and Claude. The row of engine marks on the form page shows the set that runs right now, and it can change.

- ChatGPT, Gemini, AI Overviews and AI Mode are sampled as buyers see them.
- Claude is sampled through its API and carries an API tag, because API answers can differ from the consumer app.
- Google shows an AI Overview on some searches only. The AI Overviews cell adds "Showed on X of Y searches", and its rate counts only the searches that showed one.

## Runs and confidence

The same question gets different answers from run to run, so one check proves little. Each question runs up to 50 times per engine by default. The method line at the foot of the report prints the number that actually ran.

- Every rate prints its own "x of y". The score, the platform cells, the question table and the competitor bars also show a 95% range.
- A thin bracket on each bar, and a faint band on the score ring, draw that range. Small samples make wide brackets.
- Where two brackets overlap, the gap between them could be sampling noise.
- Runs that did not return are left out, never guessed.
- A scorecard is a one-day snapshot. Rates move day to day, so read it as a snapshot, not a trend.

> **Note:** An engine cell never shows a false zero. "Not measured" means collection did not complete for that engine. "Unavailable" means a provider outage on our side: nothing was sampled and a re-run is pending. "Not shown" means AI Overviews appeared on none of the searches, so there was no answer to measure.

## How to read it

Read the page from the top. Each section answers one question.

1. **The header.** Your domain, the market, the date measured and how many questions ran across how many platforms. "How this was measured" opens the method.
2. **Questions checked.** Each question as you wrote it, with "Named in X of Y answers".
3. **Your AI visibility score.** One number out of 100: the share of measured answers that mentioned your brand, with the x of y and the 95% range beside it. At the very edges it prints below 1 or above 99 rather than a false 0 or 100.
4. **Named most and Cited most.** Beside the score: the competitor the answers name most, labeled "Named most" when it is named more often than you and "Closest competitor" when you are named as often or more. Then the page the answers cite most. A "Competitor's site" tag means that page sits on a competitor's own website, so it is not one to pitch.
5. **Key findings.** A headline and up to three findings, with "More findings" for the rest. AI writes them from the numbers on the page and nothing else.
6. **By AI platform.** One cell per engine with its share and range. With more than one question, "Show by question" opens a question by engine table. Those samples are smaller, so the per-platform figures are the steadier read.
7. **See who AI recommends instead.** You first, then competitors ranked by share. You and the competitors you entered (or that were picked from your website) are counted over the same measured answers. Other brands found in the answers are counted over all collected answers.
8. **Why each answer said what it said.** Verbatim answers, each with its engine, a "Named" or "Not named" chip, your brand and competitors highlighted, "Cites" chips that open the cited sites, and a "Why this answer" note drawn only from that answer and its sources. When no notes could be written, the section reads "What the answers actually said" and shows the excerpts alone.
9. **See which pages are influencing these answers.** The pages cited most, ranked by how many answers cite each. Different links to the same page count once, and your own site is left out. The "What next?" box suggests finding each page's editor, author or site contact.

The first 6 competitors, 6 answers and 8 cited pages are in the clear. Further rows are blurred, with a count of how many more were found. Each "Create an account" link leads to the 14-day trial sign-up.

## Share your scorecard

1. Press "Copy link" in the report header. It reads "Copied" for two seconds.
2. Paste the link anywhere. Anyone with it can view the scorecard.

- The page is not indexed by search engines, and your email never appears on it.
- A pasted link unfurls as a generic free scorecard card, with no score or brand on it.

> **Heads up:** The email says "Keep the link. It's the only copy." The link is also the only key to the page, so share it only with people you want to see the results.

## Where the command center leads

The report closes with the Autopilot band: "You've seen one day. See them all and call your next move." It previews the sourcesignal deck, with decision cards dealt from your own scorecard:

- **Cited-page pitch** and **source update:** a site your answers cite, with "Cited in X of your Y sampled answers".
- **Competitor gap:** a competitor the answers name, with its count beside yours.
- **Founder reply:** for a forum or community source, a reply you post from your own account.

Each card states a measured fact from your scorecard and what Autopilot does when you approve, such as researching the right contact and drafting the pitch. The console's status line marks it as a product preview: the console around the cards is illustrative, and its buttons send nothing.

Under the deck, the trial band takes your work email and "Start free trial": 14 days, no card needed, cancel anytime. In a trial you track the same kind of questions as an ongoing project, refreshed every week. "When the trial ends" explains that everything you tracked is kept until you choose a plan. See [Create your account](https://www.sourcesignal.ai/docs/getting-started/create-your-account.md) and [The deck](https://www.sourcesignal.ai/docs/autopilot/the-deck.md).

## Limits

| Limit | Value |
|---|---|
| Questions | 1 to 3, each 4 to 300 characters, distinct, no links |
| Brand name | 2 to 80 characters |
| Competitors | Up to 3, each 2 to 80 characters |
| Email | Work addresses only. Free webmail and disposable addresses are refused |
| Scorecards per email | One per day |
| Repeat requests | The same domain, questions and market within 14 days return the existing scorecard |

Worldwide names no country, so its questions are asked from the United States. To measure something new within 14 days, change a question or the market.

## Good to know

- On the research page, "Run the free report" carries up to three of your questions into the form, already filled in.
- When free reports are paused, the form says so and offers "Start free trial" and "Contact us" instead.
- Reply "remove" to the scorecard email to be taken off everything.
- For single checks of a page, a search or one AI answer, see the [Free tools catalog](https://www.sourcesignal.ai/docs/free-tools/free-tools-catalog.md).

## Related

- [Free tools catalog](https://www.sourcesignal.ai/docs/free-tools/free-tools-catalog.md): Every free tool, grouped by type, with what each one does, what you enter and what you get back. You don't need an account.
- [Create your account](https://www.sourcesignal.ai/docs/getting-started/create-your-account.md): Sign up with your email or a Google account, confirm your address, name your workspace and you are in. This article also covers signing in, resetting a password, the account menu and the theme toggle.
- [Trial and billing](https://www.sourcesignal.ai/docs/account-and-billing/trial-and-billing.md): Your first workspace starts with a 14-day free trial, and you need no card. This page covers what happens when the trial ends, and how to pay, change plan or cancel your subscription from Settings, then Billing.
- [AI engines](https://www.sourcesignal.ai/docs/tracking/ai-engines.md): Each project tracks its buyer questions on its own set of AI engines, chosen within your plan's count. This article covers the catalog, the AI engines card, how often each engine is checked (its cadence), paused engines and the engine status in the rail foot.
- [The deck](https://www.sourcesignal.ai/docs/autopilot/the-deck.md): Your deck deals the decisions only you can make, one card at a time. This article covers every card, the keys, undo, the daily budget and what a cleared deck means.