Built on a fixed set of customer prompts
A report is only as useful as its method. Ours starts with a set of prompts written the way your customers actually ask: plain, specific, sometimes a little impatient. That set is created during the free AI visibility audit and becomes the baseline every later month is measured against.
How the prompts are chosen
Say you run a hypothetical roofing company in Summerlin. The prompt set might include a broad question like "best roofer in Summerlin," a problem-driven one like "who can fix a leaking tile roof in west Las Vegas this week," a comparison prompt that names a competitor, and a direct check such as "what is the phone number for" your business. Broad, urgent, comparative and factual prompts reveal different things, so we include all four kinds.
Who can fix a leaking tile roof in west Las Vegas this week?
(Hypothetical answer) A few roofing companies that serve Summerlin and nearby areas are often mentioned for tile repair...
Why the prompts stay the same
If the questions change every month, the numbers stop meaning anything. Keeping the set fixed is what lets you compare this month with the first one. When a new service or area genuinely needs coverage, we add prompts as a clearly labeled addition rather than quietly swapping old ones out.
Five AI tools, run the same way
Each prompt runs across ChatGPT, Perplexity, Claude, Brave's AI answers and Microsoft Copilot. They do not all read the same web. Copilot is built on Bing, ChatGPT's search feature has drawn partly on Bing's index, and Brave Search is reportedly what Claude's web search uses. Seeing all five side by side shows whether a change is broad or limited to one tool.
What each monthly report covers
The report follows the same structure every month, so you always know where to look. Here is what each section contains.
| Report section | What it shows | Why it matters |
|---|---|---|
| Mention rate | How many prompts, per tool, named your business this month versus last month | The simplest read on whether AI answers include you at all |
| What's said | The actual wording each tool used about you, with anything inaccurate flagged | Being mentioned with the wrong phone or services is its own problem |
| Competitor share of voice | Which competitors were named, and how often, when you were not | Shows who the AI is recommending in your place |
| Sources cited | The directories, lists, threads, review sites and pages each answer linked to | Points to where the next listing or fix should go |
| Fixes shipped | Every change we made that month, with dates | Lets you connect work done to answers changed, or not changed |
| What's next | The priorities for the coming month and why | Keeps the work tied to what the answers are actually showing |
Every figure in a live report comes from prompts we actually ran that month. The table above describes the format only; it is not sample data.
Reading the report without fooling yourself
AI answers are not fixed rankings. The same prompt can produce a slightly different answer an hour later. A good report accounts for that instead of pretending it away.
Mention rate is a rate, not a ranking
There is no position one in an AI answer. What we can count is how often you are named across a stable set of prompts. That measure is less exciting than a rank, and much more honest.
Month-to-month noise is real
A single month's rise or drop can be noise. Trends over several months are the signal. We point out when a change looks meaningful and when it is probably the normal wobble of a tool that generates fresh text every time.
Why the sources section may matter most
Citations are the part of an answer you can act on. If Perplexity keeps citing a directory you are missing from, that feeds our citation and review work. If an answer quotes an outdated page, it becomes a job for correcting wrong AI answers.
How the report drives next month's work
The report is not a scorecard we hand over and forget. Each section feeds a specific part of the Superintelligent Optimization program. Inaccurate wording often traces back to identity signals our entity and schema fixes address. Thin or vague descriptions of your services point to answer content an AI can quote. A tool that never cites your site at all is a reason to recheck whether AI crawlers can reach your pages.
Fixes shipped, in plain language
Every report lists what we changed that month and where, written so a busy owner can scan it in a minute. If nothing shipped in a given area, the report says so.
What's next, and what we cannot promise
We close each report with the coming month's priorities. What we will never include is a promise that a fix will show up in a given tool by a given date. The AI companies decide when their answers change; we improve the inputs and report the outputs honestly.
Want to see your baseline before anything else? Book a free AI visibility audit, or call (725) 525-8508 to ask how the monthly reports work.
Common questions
Which AI tools are included in the monthly reports?
ChatGPT, Perplexity, Claude, Brave's AI answers and Microsoft Copilot. Every prompt in your set is run across all five each month.
Will the report show our ranking in ChatGPT?
AI answers do not have rankings in the traditional sense. The report shows how often you are named across a fixed prompt set, what is said about you, and which competitors appear when you do not.
Can we add prompts after the program starts?
Yes. New prompts are added as a labeled addition so the original set stays comparable month to month.
How is the report delivered?
As a written monthly report sent to you, with the same sections every month. We are also happy to walk through it with you by phone.