---
title: "Your AI Visibility Score Is an Impressions Deck With a New Haircut"
description: "AI visibility scores can show where your brand is missing. Alexander Perelman argues that a monthly gap report is no substitute for fixing the page and measuring what changed."
url: "https://groas.com/post/your-ai-visibility-score-is-just-impress"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/5e93af4d-6e5f-46c4-9c48-1d085dbfe05e.png"
published: "2026-10-13T05:27:25.898Z"
modified: "2026-10-13T05:27:25.968Z"
---

October 13, 2026 · 9 min read

# Your AI Visibility Score Is an Impressions Deck With a New Haircut

[Alexander PerelmanHead Of Product @ groas](https://groas.com/author/alexander-perelman)[LinkedIn](https://www.linkedin.com/in/alexander-433793253/)

![An old, coffee-stained impressions report sits in a barber chair sporting a slick new haircut, mocking AI visibility scores as rebranded old metrics.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/5e93af4d-6e5f-46c4-9c48-1d085dbfe05e.png)

In this article

1. [A visibility score shows presence, not progress](#a-visibility-score-shows-presence-not-progress)
2. [The prompt meter runs even when nothing gets fixed](#the-prompt-meter-runs-even-when-nothing-gets-fixed)
3. [“Content gap identified.” Great. Who’s writing it?](#content-gap-identified-great-whos-writing-it)
4. [Reporting scales; repair is the part that hurts](#reporting-scales-repair-is-the-part-that-hurts)
5. [Check this before you buy another score](#check-this-before-you-buy-another-score)
6. [A gap list without priority is just a longer gap list](#a-gap-list-without-priority-is-just-a-longer-gap-list)
7. [Put the work before the scoreboard](#put-the-work-before-the-scoreboard)

Most AI visibility dashboards have rebuilt the worst habit of 2015 PPC reporting: charge monthly for a pretty number, then leave the client to do the fixing. The score goes up, the score goes down, and nobody can tie the movement to revenue.

I know the old version because I used to send impressions decks. Impressions up. CTR down a touch. Big logo on page one. The client still had to rewrite the landing page. I told myself the reporting was the work. It was not. Now a founder can pay to learn their brand scored 32 while a competitor scored 68, then receive the same homework: go fix your content. **A report that names a gap and does not close it is an invoice, not a service.**

## A visibility score shows presence, not progress

Here is the mechanism vendors tend to hurry past. An [AI visibility score measures overall presence across AI platforms](http://searchengineland.com/tools/ai-visibility-checker): how often your brand appears, gets cited, or gets mentioned. One vendor example puts a brand at 32 and a competitor at 68. [Share of voice is calculated across the prompts you track, factoring in answer appearances, citations, and mentions](http://searchengineland.com/tools/ai-visibility-checker). The pitch follows neatly: identify the gap, study the competitor’s content, then create or refine yours until you appear.

That last step is doing a lot of lifting. It is also the step the dashboard leaves on your desk.

I used to sell the PPC version of this logic. Impressions meant an ad rendered. They did not mean someone read it, cared, or bought anything. Even the underlying search numbers need care. In [Search Console reporting for generative AI](https://www.searchenginejournal.com/google-admits-search-console-reporting-for-ai-search-is-inadequate/589236/), an AI Overview impression can count when the block renders whether the user saw its link; links behind Show More are treated differently until expanded, and links inherit the position of the whole block. As [one analysis of AI Overview impressions](https://www.playwire.com/blog/ai-overviews-impressions-dont-mean-what-you-think) argues, those counts are directional at best. They do not belong in a revenue forecast because a line moved in the right direction.

A score of 32 can tell you something useful: across a chosen set of prompts, on a given day, a model mentioned your brand less often than another. It cannot, by itself, tell you which prompt brings customers, which page should have earned a citation, what kept that page out, or whether a rewrite brought in business. Those are different questions. If the score is all you have, you have not answered them.

Then there is the raw reporting beneath the glossy line. [Search Console’s new AI data, as described here, reports impressions per URL with country and device but no filter for the feature that produced them and no click data](https://www.practicalecommerce.com/search-consoles-new-ai-data-is-a-start). A URL can register an appearance even when the link is tucked behind Show More. That is useful context if you know its limits. Put it in a dashboard, call it momentum, and it becomes the old impressions deck with a new haircut. **Presence is a starting observation, not a business result.**

## The prompt meter runs even when nothing gets fixed

The next screen is the upsell. Start with 15 prompts. Then try to cover your products, competitors, and cities without leaving out the searches you care about. Suddenly the small plan is not so small. [Otterly lists Lite at $29 a month for 15 prompts, Standard at $189 for 100, and Premium at $489 for 400](https://otterly.ai/pricing), with Claude, Gemini, and AI Mode as paid add-ons. A [comparison of monitoring tools](http://ziptie.ai/blog/best-generative-ai-search-monitoring-tools/) puts Otterly’s 100-prompt plan beside Profound’s $499 Lite tier and ziptie’s $69 plan for 1,500 queries. Different meters, same question for the buyer: what happens after the query is counted?

I used to tell clients more keywords meant more control. I was wrong. More keywords could also mean more structure to maintain and thinner data in every ad group. Here, more prompts can mean more rows to inspect without a better reason to change any particular page. The subscription keeps running while someone on your team works out what the chart is asking them to do.

[Many visibility tools are monitoring-first](http://ziptie.ai/blog/best-generative-ai-search-monitoring-tools/). They show mentions and charts, but leave optimization to the buyer; the same comparison describes Otterly as monitoring-only, with weekly refreshes and no optimization guidance. Monitoring can be useful. Selling it as though the work is handled is something else. Say you spend $20k a month on search. You would not hire a manager solely to send a screenshot of CPA rising. You would expect them to bring CPA down and explain what they changed. **Why pay per prompt to watch yourself lose?**

![Ink cartoon of a marketer paying monthly for a chart that says you lost](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/9528c11a-c13b-4c42-87e5-68c8cc2cc81e.png)

## “Content gap identified.” Great. Who’s writing it?

This is the screen that gets me. A red row says _content gap identified_: the competitor was cited; you were not. Under it sits a suggestion box. Rewrite the page. Add an FAQ. Earn a citation. What the deck calls _actionable insight_, I call homework with a subscription fee attached.

I know the PPC version. I could flag a falling Quality Score, recommend a new landing page, and write up the audit. The client still needed someone to write the copy and ship the page. My report described the hole. It did not fill it. A dashboard can make that handoff look smoother, but the handoff is still there. Someone has to decide which page matters, make the change, publish it, and check whether the answer changed afterward.

There is nothing wrong with buying a task list if you want a task list. Just do not confuse it with coverage. **If the tool stops at the gap, the work starts when its report ends.**

## Reporting scales; repair is the part that hurts

A dashboard that checks a set of prompts once a week has a repeatable job. Closing a gap demands a messier chain of work: connect the prompt to a page, decide what the page is missing, change it, publish it, test the prompt again, and record what happened. One process produces another chart. The other might change what the next answer says.

That difference explains why the market is crowded with reports. The 2015 agency version was familiar to anyone who lived through it: review the account, send the PDF, charge for management. Rebuilding ad groups, writing ads, and fixing tracking took more work than describing what was wrong. Reporting scaled. Repair did not. A monthly visibility dashboard can recreate the same arrangement with better typefaces and a smaller invoice.

If you run a small business and want to monitor your brand without hiring an agency, you may need fewer prompts, not more. Track the ones tied to pages you can change. Ten buying prompts with citation history are more useful than 400 vanity prompts with a single score if you have to decide what to do on Monday. That is not an argument against measurement. It is an argument against measuring so broadly that nobody owns the next move.

Ask a vendor to show you the prompt, the answer, the cited source, and the page on your site that should have competed. If the trail ends at “competitor mentioned,” you have been informed. **Information is not the bottleneck when execution remains untouched.**

## Check this before you buy another score

If you still want software to automate AI search visibility reporting, I get it. A clear record of where you appear can help, provided it leads to a change you can inspect. Use this checklist before paying for another month. I would rather see a tool do a few of these jobs properly than track hundreds of prompts and leave every decision to you.

1. Show the exact prompt, full AI answer, and cited source behind each mention. A rolled-up score is not a receipt.
2. **Map each buying prompt to the page on your site that should have earned the citation.** A competitor URL alone does not identify your fix.
3. **Ship the fix, not a PDF about it.** Rewrite and publish the page rather than handing you instructions to do both.
4. Log each change with its date, reason, and what it replaced. Otherwise, the next result has no usable history.
5. **Show movement separately in ChatGPT, Gemini, and Google AI Overviews.** This is the item most often skipped; without that history, you can credit the wrong change, undo the right one, and pay another month to watch the score wobble.

## A gap list without priority is just a longer gap list

Ask which platform shows content gaps preventing citation and you will find no shortage of raised hands. The output tends to look familiar: a table of prompts where a competitor was cited and you were not, followed by suggestions to add a definition, a statistic, or an FAQ block. That tells a business owner where to look. It does not settle which row deserves work first.

Which of those prompts leads toward a call, a demo, or a checkout? Which page should have won the citation? What, specifically, would you change on that page? Without that connection, adding rows makes the report longer, not more useful. I have made PPC audits that suffered from the same problem: plenty of wasted spend identified, no fix attached. A long diagnosis can feel thorough right up until someone asks who is doing the work.

For a closer look before buying, there is a [ranking of AI search visibility tools by whether they ship fixes or stop at dashboards](https://groas.com/post/ai-visibility-tools-ranked-by-what-they). If the prompt meter itself is the question, read [whether tracking AI prompts is worth paying for](https://groas.com/post/is-tracking-ai-prompts-worth-paying-for). My short version is less polite than either sales deck: tracking earns its keep when a prompt points to a page, an owner, and a change.

Groas, the shop I write for, takes the other half of that job seriously. It maps prompts to pages, publishes the rewrite, and logs what moved in ChatGPT, Gemini, and AI Overviews. That is the better alternative to monitoring-first software for a business that wants the gap closed, not merely circled in red. **A tool that cannot change your site cannot do the work your visibility score is asking for.**

## Put the work before the scoreboard

The useful PPC report was the one a client could read in a minute: spend, CPA, what changed, why, and what happens next. Every number belonged to a decision. The scoreboard followed the work instead of pretending to be the work.

AI visibility needs that order. Track a small set of prompts with buying intent. Tie each to the page that should earn the citation. Make the change and publish it. Test the same prompt again in ChatGPT, Gemini, and AI Overviews. Record what moved and what you replaced. Then decide whether the change deserves to stay. That is a loop someone can manage. A score floating up and down while the site stays untouched is a screenshot.

Vendors will say the score is only the starting point. Fine. Impressions were a starting point too. The trouble begins when a starting point is packaged as the monthly service and the buyer is left to supply the ending. I believed for too long that a clear report proved good account management. It proved I could describe an account. The distinction mattered as soon as someone needed the landing page fixed.

Cancel the dashboard that charges you to watch. Keep the system that changes the page, checks the answer, and shows you what happened. **A gap report that will not close the gap is an invoice with a graph.**

## Frequently Asked Questions

### What does an AI visibility score actually measure?

An AI visibility score measures how often your brand appears across AI platforms, including answer appearances, citations, and mentions across the prompts you track. It shows presence on a given day, but it does not tell you which prompts bring customers, which page should have earned a citation, or whether a change produced business results.

### Are AI Overview impressions in Search Console reliable?

They are directional at best. An impression can count when the block renders whether the user saw your link, links behind Show More are treated differently until expanded, and links inherit the position of the whole block. Because of that, these counts do not belong in a revenue forecast.

### Do AI visibility tools charge more as you add prompts to track?

Yes, most meter by prompt count. Otterly lists Lite at $29 a month for 15 prompts, Standard at $189 for 100, and Premium at $489 for 400, with Claude, Gemini, and AI Mode as paid add-ons, while Profound's Lite tier costs $499 and ziptie charges $69 for 1,500 queries. More prompts can mean more rows to inspect without a better reason to change any particular page.

### Do AI visibility tools actually fix anything for you?

Many are monitoring-first. They show mentions and charts but leave optimization to the buyer; a comparison of monitoring tools describes Otterly as monitoring-only, with weekly refreshes and no optimization guidance. If the tool stops at the gap, the work starts when its report ends.

### A tool flagged a content gap against a competitor. What should I do with that?

Work out which of those prompts leads toward a call, demo, or checkout, which page should have won the citation, and what specifically to change on that page. Without that connection, adding rows makes the report longer, not more useful, and the gap stays open.

### How many AI prompts should a small business track?

You may need fewer prompts, not more. Track the ones tied to pages you can change, because ten buying prompts with citation history are more useful than 400 vanity prompts with a single score when you have to decide what to do on Monday.

### What should I check before paying for an AI visibility tool?

Check that it shows the exact prompt, full AI answer, and cited source behind each mention; maps each buying prompt to the page on your site that should have earned the citation; ships the fix rather than a PDF about it; logs each change with its date, reason, and what it replaced; and shows movement separately in ChatGPT, Gemini, and Google AI Overviews.

### What does closing an AI visibility gap actually involve?

Closing a gap demands connecting the prompt to a page, deciding what the page is missing, changing it, publishing it, testing the prompt again, and recording what happened. A dashboard that only checks a set of prompts once a week produces another chart; the repair is the part that actually changes what the next answer says.

## Pay For Results, Not For Hours

Businesses buy the outcome, agencies resell it, and groas answers for it either way.

[See If You Qualify](https://groas.typeform.com/to/xC1bQNUT)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#webpage","url":"https://groas.com/post/your-ai-visibility-score-is-just-impress","name":"Your AI Visibility Score Is an Impressions Deck With a New Haircut","description":"AI visibility scores can show where your brand is missing. Alexander Perelman argues that a monthly gap report is no substitute for fixing the page and measuring what changed.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/5e93af4d-6e5f-46c4-9c48-1d085dbfe05e.png"},"breadcrumb":{"@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#breadcrumb"},"dateModified":"2026-10-13T05:27:25.968Z","mainEntity":{"@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"Your AI Visibility Score Is an Impressions Deck With a New Haircut","item":"https://groas.com/post/your-ai-visibility-score-is-just-impress"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#article","url":"https://groas.com/post/your-ai-visibility-score-is-just-impress","mainEntityOfPage":{"@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"Your AI Visibility Score Is an Impressions Deck With a New Haircut","description":"AI visibility scores can show where your brand is missing. Alexander Perelman argues that a monthly gap report is no substitute for fixing the page and measuring what changed.","datePublished":"2026-10-13T05:27:25.898Z","dateModified":"2026-10-13T05:27:25.968Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/5e93af4d-6e5f-46c4-9c48-1d085dbfe05e.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"page, score, visibility, prompts, impressions, gap, visibility score, change","wordCount":1923,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"Ex Goldman Sachs and Ex Stanford Computer Science","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/your-ai-visibility-score-is-just-impress#faq","mainEntity":[{"@type":"Question","name":"What does an AI visibility score actually measure?","acceptedAnswer":{"@type":"Answer","text":"An AI visibility score measures how often your brand appears across AI platforms, including answer appearances, citations, and mentions across the prompts you track. It shows presence on a given day, but it does not tell you which prompts bring customers, which page should have earned a citation, or whether a change produced business results."}},{"@type":"Question","name":"Are AI Overview impressions in Search Console reliable?","acceptedAnswer":{"@type":"Answer","text":"They are directional at best. An impression can count when the block renders whether the user saw your link, links behind Show More are treated differently until expanded, and links inherit the position of the whole block. Because of that, these counts do not belong in a revenue forecast."}},{"@type":"Question","name":"Do AI visibility tools charge more as you add prompts to track?","acceptedAnswer":{"@type":"Answer","text":"Yes, most meter by prompt count. Otterly lists Lite at $29 a month for 15 prompts, Standard at $189 for 100, and Premium at $489 for 400, with Claude, Gemini, and AI Mode as paid add-ons, while Profound's Lite tier costs $499 and ziptie charges $69 for 1,500 queries. More prompts can mean more rows to inspect without a better reason to change any particular page."}},{"@type":"Question","name":"Do AI visibility tools actually fix anything for you?","acceptedAnswer":{"@type":"Answer","text":"Many are monitoring-first. They show mentions and charts but leave optimization to the buyer; a comparison of monitoring tools describes Otterly as monitoring-only, with weekly refreshes and no optimization guidance. If the tool stops at the gap, the work starts when its report ends."}},{"@type":"Question","name":"A tool flagged a content gap against a competitor. What should I do with that?","acceptedAnswer":{"@type":"Answer","text":"Work out which of those prompts leads toward a call, demo, or checkout, which page should have won the citation, and what specifically to change on that page. Without that connection, adding rows makes the report longer, not more useful, and the gap stays open."}},{"@type":"Question","name":"How many AI prompts should a small business track?","acceptedAnswer":{"@type":"Answer","text":"You may need fewer prompts, not more. Track the ones tied to pages you can change, because ten buying prompts with citation history are more useful than 400 vanity prompts with a single score when you have to decide what to do on Monday."}},{"@type":"Question","name":"What should I check before paying for an AI visibility tool?","acceptedAnswer":{"@type":"Answer","text":"Check that it shows the exact prompt, full AI answer, and cited source behind each mention; maps each buying prompt to the page on your site that should have earned the citation; ships the fix rather than a PDF about it; logs each change with its date, reason, and what it replaced; and shows movement separately in ChatGPT, Gemini, and Google AI Overviews."}},{"@type":"Question","name":"What does closing an AI visibility gap actually involve?","acceptedAnswer":{"@type":"Answer","text":"Closing a gap demands connecting the prompt to a page, deciding what the page is missing, changing it, publishing it, testing the prompt again, and recording what happened. A dashboard that only checks a set of prompts once a week produces another chart; the repair is the part that actually changes what the next answer says."}}]}]}
```