Free tool · Agency pitch

Your client’s buyers now ask an AI engine about a brand before they visit its site, and the answer they get is written from pages the brand doesn’t control. The free AI Visibility Report shows you that answer in about four minutes: what the engine says, which pages it pulled from, and which fixes to make first, with your agency’s name on it.

~4 min
from domain to report
$0
no card, no account
5 engines
one engine per report
Your logo
on the report and PDF

How to run the report, step by step

You need a domain and about four minutes. There is no account and no card.

Step 1 of 4

Enter the brand, the engine and the country

Type the client’s or prospect’s domain. Choose the country the engine should be asked from, so the answers match what buyers there actually see. Then pick one engine: ChatGPT, Perplexity, Gemini, Google AI Mode or Google AI Overviews. Each report covers one engine. If you want to compare two, run two reports.

AI Visibility Report input form filled in with the domain apple.com, United States as the country and ChatGPT as the AI engine

Step 2 of 4

This part is optional, but it’s the point of the tool. Add your agency name and a logo (PNG, JPEG, WebP or SVG, up to 512 KB). They appear on the report and on the PDF, and the closing section carries your booking link.

Agency details section of the AI Visibility Report form with an agency name field and a logo upload button

Step 3 of 4

Run it and let it work

Click Run my free audit. The engine is asked about the brand repeatedly, each time in a fresh session, all inside one short window. Every answer and every citation is saved. Then the tool reads a handful of the brand’s own pages to see what the site says. The report opens in your browser when it’s ready. It isn’t published online and the brand isn’t notified.

The Run my free audit button changing to Queuing your audit while the AI Visibility Report starts

Step 4 of 4

Open the report and download the PDF

Read the results in the browser, then download the PDF to take to your prospect. A prospect who opens it sees a preview and unlocks the full report with a work email.

Finished AI Visibility Report for Apple open in the browser with the section tabs across the top

The report opens in your browser with your agency name on it.

Work email prompt that unlocks the full AI Visibility Report, shown over the blurred report

The preview is free. A work email unlocks the full report.

Good to know: you get one audit per brand, engine and country every 30 days. Run it again inside that window and you’ll be shown the existing report.

How to read the results

Every figure is counted from the saved answers and the brand’s own site. The written commentary explains those numbers; it doesn’t add new ones. The example below is a real run on apple.com with ChatGPT in the United States. The report has seven numbered sections and a closing page.

Section 01

Why this matters now

Sets the scene with cited market statistics. These are third-party figures, not measurements of the brand, so use them to open the conversation.

In the Apple example: 43% of US online shoppers used an AI assistant for product research in the last 90 days (MarTech, 2026).

Section 02

Where you lose the sale

Maps the buyer’s path and marks the steps where the engine’s answer can cost the brand the sale.

In the Apple example: the problem isn’t awareness, it’s trust at the moment of purchase. A buyer who hears something that gives them pause sticks with the name they already know, and the brand never sees that person.

Section 03

What ChatGPT says

Splits what the engine says into what is working and what is costing the brand, with how often each fact appears. A fact in nearly every answer is what buyers read. A fact that never shows up is a gap.

In the Apple example: the headquarters appeared every time, iPhone 18 Pro by name in about 80% of answers, Apple Watch Series 12 in about 75%, and the founding year or founder in about 70%. The customer count, the price, and any rating or award never appeared.

What ChatGPT says section of the Apple report showing which facts appear in the answers and which never do

Section 04

Where it looks

Shows which pages feed the answers, split between the brand’s own site and everyone else’s. A client’s best page sitting uncited is a ready-made scope of work.

In the Apple example: 59% of what ChatGPT drew on came from apple.com and 41% from other sites. Two newsroom pages each appeared in about a third of answers, and sec.gov appeared in about 70%.

Where it looks section of the Apple report splitting sources between apple.com and other sites

Section 05

Why

Names the root causes behind the gaps, in order of importance, so the fix list has a reason behind every line.

In the Apple example: the biggest reason is small, mixed signals on its own site (weak organisation schema, no clear definition, conflicting facts). The second is that nobody outside the site tells the company’s story, because no third-party page defines it.

Section 06

The goal

Puts the answer the engine gives today beside the answer to work toward, so the client sees the gap at a glance.

In the Apple example: today’s answer opens with a generic line, “a major American technology company headquartered in Cupertino, California”, and lists product categories instead of current products. The founding year reaches about 70% of answers, the hardware-and-design differentiator about 40%, and the 2.5 billion installed base none. The target answer carries the founding date, the headquarters, the current product line, the installed base and the differentiator, each backed by proof on the brand’s own pages.

Section 07

The plan

A timeline of fixes, quickest wins first. Website changes show up first; outside coverage takes longer because AI has to find and trust new sources.

In the Apple example: Weeks 1 to 2, fix the website. Write a definition of what the company does, at the very top of the homepage. Choose one founding year. Make sure every page shows that same founding year. Finish filling in the Organization schema. Finally publish a worked example on a page that the engine already cites so that it has something to pull from.

Weeks 4 to 12, get others talking. Get three third-party pages (a Crunchbase profile, the LinkedIn company page and an industry directory) to define the company, not just mention it.

Week 12, re-check. Ask the same question again and measure the change against today’s figures.

The plan section of the Apple report with weeks one to two fixes for the website

The report closes with a next-step page that carries your booking link and the case for an ongoing retainer.

What to do next

1Start with the stuff the brand controls. Look for checks that failed because the site says one thing and the engine says another. Those are the cheapest wins on the list, and the report hands you a sentence the client can paste onto their site today.
2Then go after the pages the engine actually cites, not only the brand’s own. Say it keeps pulling from directories, review sites, or press that’s years old. That means the problem lives off the site, so you’ll want to plan some outreach and get the brand described properly in the places the engine already reads.
3Take the report into the pitch. Open with what the engine told this brand’s buyers, then walk through the ranked fixes. The final section makes the case for an ongoing GEO/AEO retainer.

Where the full Wellows platform goes further

The report is a one-time sample: one engine, one window, one brand. It tells you where a brand stands today. The Wellows platform does the ongoing work: it tracks citations daily, shows which cited pages mention the brand and which don’t, and gives you the outreach and content moves to close the gap. Run the report to open the conversation, then use Wellows to do the work.

Run it on your next prospect

Try the free AI Visibility Report, then see how Wellows helps agencies turn the findings into retainers.

Try the free AI Visibility ReportWellows for agencies

Each report is an automated sample of one engine in one window, not an evaluation of the company.


Yes. It runs online with no credit card and no account required. Your prospect sees an initial preview and unlocks the complete report using a work email address.

Enter the brand’s domain, select a target country and one AI engine, and run the report. In approximately four minutes, you will see how the engine describes the brand, which source pages it references, and where the brand’s own site has content gaps.

ChatGPT, Perplexity, Gemini, Google AI Mode, and Google AI Overviews. Each report evaluates a single engine, meaning a Perplexity audit and a ChatGPT audit are generated as separate runs.

Yes. Your agency’s name and logo appear directly on the web report and the downloadable PDF, while the final section includes your booking link and contact details for client follow-ups.