Get one AEO fix in your inbox every week. Join the newsletter →

Video · about 2 minutes

What does an AI visibility report show? How to read it.

Read the prompts before the numbers. An AI visibility report submits a set of questions to AI tools and records what comes back. Most reports show four numbers: mention rate, citation rate, recommendation rate, and share of voice. It is a useful snapshot, not a live measurement. Treat it as directional, not definitive.

What a report measures

A sample, not a live feed.

An AI visibility report measures how your business appears in the answers AI tools like ChatGPT, Perplexity, Claude, Gemini, and Google's AI search features give. It is not a traditional SEO ranking report, and the two should not be read the same way.

What it should tell you

The method behind the numbers.

Which questions were tested. These are called prompts. A good report lists them so you can judge whether they match what your customers ask.
Which AI platforms were included. A report covering only one platform shows a partial picture.
How many answers were analyzed. A report based on 10 responses is less reliable than one based on 100.
The time period covered. AI answers change as models update and new information appears online.

If a report does not explain its methodology, that is useful information too. You cannot assess how much weight to give the numbers.

What it cannot capture

Every answer a customer sees.

AI answers vary by location, by device, by whether web search is active, and by how the question is phrased. Google has noted directly that no third-party tool has access to its internal ranking or AI systems.

A report that shows your business in 40% of tested answers does not mean 40% of customers will see your name. It means your business appeared in 40% of the specific answers the tool collected.

Directional, not definitive

Want to be the business ChatGPT names?

See the VIP programsociallysmashing.com/recommended/vip

The four numbers

What each one means, and what it doesn't.

Most AI visibility reports organize their findings around four core metrics.

Mention rate

Were you named at all?

The percentage of tested answers in which your name appeared, as a recommendation, a passing reference, a comparison, or a footnote. It tells you whether AI tools know your business exists in the context of the questions asked. It does not tell you how you were described.

What to check: the actual answers. A business mentioned as a cautionary example has a high mention rate and a problem.

Citation rate

Was your site a source?

A citation is when an AI answer links to or clearly draws from your website. In Bodhi & Co's 2026 AI Visibility Report, businesses' own websites made up only 78 of the 1,685 source citations logged in its September tests, roughly 5%. AI tools frequently cited directories, review platforms, and third-party publications instead.

What to check: which sources the AI cited instead of you, and whether they mention you or only competitors.

Recommendation rate

Were you suggested?

A recommendation is when an AI answer explicitly suggests your business as an option to choose or contact. This is the metric most directly connected to customer acquisition. EMARKETER research published in October 2026 found 76% of the marketers surveyed by EMARKETER and Semify consider appearing as a citation or recommendation in AI-generated responses extremely or very important.

What to check: the prompts where you are recommended. They show what AI associates your business with most strongly.

Share of voice

How do you compare?

Your mentions or citations as a percentage of the total across all businesses tracked in the same report. It is a competitive metric, not an absolute one.

If the report tracked three local competitors and you have a 60% share of voice, that tells you something useful. If it tracked only one minor competitor, the number is less informative.

Metric	What it measures	What it does not measure
Mention rate	Whether your name appears in AI answers	How you were described or whether the mention helped
Citation rate	Whether your website was used as a source	Whether that citation led to a recommendation
Recommendation rate	Whether AI explicitly suggested your business	Whether that suggestion reached a real customer
Share of voice	Your presence relative to tracked competitors	Businesses outside the tracked set

Before you act

Three questions that make a report useful.

Do the prompts match what my customers ask?

The prompts determine everything. Too generic ("What is a good plumber?") or too narrow ("Who installs copper pipes in downtown Austin?") and the results will not reflect your customers. Read the prompt list before the numbers. If it does not fit, note the gap and ask for a revised prompt set.

Are the numbers improving or declining?

A single report is a snapshot. Moz's October 2026 AI visibility workflow treats prompt results as directional and advises comparing the same prompts over time. Treat your first report as a baseline. Movement, in either direction, is more actionable than the first number.

Does it connect to an outcome I can verify?

AI visibility metrics are upstream of business outcomes. When your recommendation rate improves on important prompts, does branded search, direct traffic, or inquiry volume move the same way? If visibility improves and none of those move, something in the chain is broken.

A useful check: Pull your Google Analytics data and look for sessions arriving from AI referral sources. Google Search Console now includes a Generative AI performance report that shows impressions for your pages in AI Overviews and AI Mode. These are not complete pictures, but they give you a reference point outside the visibility report itself.

Common misreads

Four ways reports get read wrong.

Misread 1

A high mention rate means good standing.

A mention confirms AI tools associate you with the topic. It does not confirm the association is positive, prominent, or commercially useful. Read a sample of the actual answers. Are you a recommended option, one of several alternatives, or a background reference? The position and framing matter more than the count.

Misread 2

A low score means you are failing.

Bodhi & Co found that across 111 business websites it audited in 2026, the average AI visibility score was 32.6 out of 100, compared to 70.5 for traditional SEO, both graded by Bodhi & Co's own audit. AI visibility was the lower score on every single site tested. Low scores are common, not exceptional.

Misread 3

Benchmarks without context.

A benchmark is only useful if the group is comparable: same industry, same geographic scope, same customer questions, same AI platforms. A local service business should not measure itself against a national e-commerce brand. Ask how the benchmark was built before using it.

Misread 4

Expecting the report to explain why.

Reports record what happened. If you are not recommended for a type of question, the report shows the gap, but it cannot tell you whether the cause is thin content, missing third-party mentions, crawling issues, or something else. The report is the starting point for an investigation, not the conclusion of one.

After you read it

Five steps, in order.

Step 1

Check the prompts.

Confirm the questions tested are relevant to how your customers search. If they are not, that is the first thing to fix.

Your check: your customers' words

Step 2

Find your recommendation gaps.

Sort the prompts by recommendation rate, lowest to highest. The prompts where you are not recommended, especially where competitors are, are your most specific starting points.

Your check: competitor named, not you

Step 3

Check what sources the AI cited.

When you did not appear, which sources did? Directories, review platforms, or publications that cover your category show where your business needs to be present and described accurately.

Step 4

Let AI crawlers in.

Some sites have settings that block AI tools from reading their pages. Check that your site allows GPTBot, ClaudeBot, PerplexityBot, and Google-Extended. A blocked site will not appear in answers regardless of how good the content is.

Step 5

Set a date to re-run.

One report is a baseline. Re-run the same prompts under the same conditions in 30 to 60 days and compare.

Your check: re-run date set

The honest part

One input, not the whole picture.

The measurement issue is real. EMARKETER research from October 2026 found that 56% of the marketers surveyed cite measuring ROI from AI visibility as a challenge, the most-cited one. That difficulty does not make the report useless. It means the report works best as one input alongside the business signals you already track: inquiry volume, branded search, direct traffic, and what new customers tell you about how they found you.

Your customer questions are the starting point for all of this. If the prompts in a report reflect those questions, the report gives you a structured way to see whether AI tools are sending the right people toward your business, and where that connection is still missing.

VIP

Want to be the business ChatGPT names?

Have Nicole Jolie do this for you. She runs the prompts, reads the reports, and writes the pages in your buyers' questions. You answer questions and approve pages.

See the VIP programsociallysmashing.com/recommended/vip

One fix a week

Get one AEO fix in your inbox every week.

Short, practical, and built to be used the same day.

The newsletter

One AEO fix each week, plus first access when the community opens.

Join the newsletter

YouTube

Short videos, each answering one question about being found by AI.

Subscribe on YouTube

The conversation in between.

Follow me on LinkedIn

Questions

AI visibility reports, answered.

What does an AI visibility report measure?

It submits a set of questions, called prompts, to AI tools and records what comes back. It then analyzes those answers for your business name, your website, and how you were described. A report is a sample taken at a specific time, using a specific set of questions, on specific platforms. It is a useful snapshot, not a live measurement.

What is the difference between a mention, a citation, and a recommendation?

A mention is when your business name appears anywhere in an AI answer. A citation is when an AI answer links to or clearly draws from your website as a source. A recommendation is when an AI answer explicitly suggests your business as an option to choose or contact. Being mentioned or cited is not the same as being recommended.

Does a low AI visibility score mean my business is failing?

No. Low scores are common, not exceptional. Bodhi & Co's 2026 report found that across 111 business websites it audited, the average AI visibility score was 32.6 out of 100, compared to 70.5 for traditional SEO, both graded by Bodhi & Co's own audit. The more useful question is whether your score is improving.

What should I check first in an AI visibility report?

The prompts. Confirm the questions the report tested match what your customers actually ask. If they do not, the report is measuring a situation that does not apply to your business. Note the gap and ask for a revised prompt set before drawing conclusions.

How often should I re-run an AI visibility report?

Treat your first report as a baseline. Run the same prompts on the same platforms in 30 to 60 days and compare. That gives enough time for content changes to register, without waiting so long that you cannot connect the changes to the results.