？
It doesn't know you
Ask about your category and your name never comes up. You're not in the conversation at all.
When your buyers ask ChatGPT, Claude, Gemini, or Perplexity for a recommendation, do you come up — and is what they say even right? The AI Visibility Audit shows you where you stand and how to climb. It's included free with a Workflow Margin Scan — or $1,500 on its own.
This is the outside-in view — how AI sees and recommends you to buyers. (Our Workflow Margin Scan is the inside-out one — it unlocks the margin trapped in your operations.) Pick the depth that fits.
A one-page read of what one AI engine says about you today, the three questions AI asks — does it know you, trust you, bring you up — and three fixes you can start now. Your $99 is credited in full toward the full audit if you upgrade within 10 business days.
See the SnapshotAll five AI surfaces — ChatGPT, Claude, Gemini, Perplexity, and Google's AI answers — the complete diagnosis across understandability, credibility and deliverability, and your prioritized 90-day plan. Then $675/quarter to track and work the changes — and the AI Content Roadmap when you're ready to build.
See the full AuditMore and more people ask an assistant "who's good at X?" before they ever open a search engine or land on your site. If the AI doesn't know you, gets your facts wrong, or names a competitor instead — you lose the deal before you knew it existed. Most businesses haven't touched this yet. That's the opening.
Ask about your category and your name never comes up. You're not in the conversation at all.
AI describes you with stale, vague, or flatly incorrect facts — or confuses you with a similarly-named company.
For the exact questions you should win, the assistant confidently recommends someone else.
AI recommends a business only when three things are true. The scan looks at all three.
Does AI actually know who you are, what you do, and where — and are the facts right? If it's confused, nothing else matters.
Does AI trust you enough to recommend you outright — or does it hedge with "appears to" and "claims to"? Trust is earned across sources.
Do you actually show up when someone asks the questions that should surface you — not just when they search your name?
Already fixing your workflows? It's on us. Want it on its own? It's priced to be an easy yes.
If you're already booking a Workflow Margin Scan, the AI Visibility Audit comes with it — same engagement, no add-on fee. We look inward and outward at once, and you get the full picture of your business.
See the Margin ScanA full Understandability / Credibility / Deliverability diagnosis and a plan to climb. $1,500 for the first audit — then $675 a quarter: a fresh full measurement, a new report, and a review of it with you.
See a sample & buyEvery audit comes with drop-in assets — schema, an llms.txt file, and entity copy — that help AI engines understand you correctly. We'll confirm scope before anything starts.
We scan businesses two directions. The Workflow Margin Scan looks inward — where manual work is draining your margin. The AI Visibility Scan looks outward — how the market, and the AI that increasingly mediates it, sees you.
Different problems, same habit: find the truth first, then fix what's worth fixing. Plenty of businesses need both.
See the Workflow Margin ScanAI visibility got crowded fast. Most of what's for sale falls into two buckets — and neither one tells you what to do on Monday.
A subscription that tracks mentions over time. Useful if you have a team to act on it. You're renting a number, and the number doesn't fix anything — and it stops the day you stop paying.
Monitoring has gotten cheap — around a cent and a half per check at volume, with entry plans from $50 a month. That's a good thing, and it's why we don't sell one. When the measurement is a commodity, the part worth paying for is knowing what it means. (Published pricing across eight tools, September 2026.)
A quick audit that returns a grade out of 100. Fast and cheap. But most run each question once — and AI answers change run to run, so a single pass is closer to a coin flip than a measurement.
Every question asked several times across five engines, reported as a rate. Why you get missed, not just that you do. Every claim traced to a saved transcript. And a 90-day plan you can hand to whoever does the work.
We publish our full method, ungated — read exactly how the audit is run. If you'd rather run it yourself with our method, that's a fair outcome too.
An AI visibility report can be produced in ten minutes or over several days, and once it is formatted it is hard to tell which one you are holding. These four checks take about a minute and work on any report from any provider.
Not how many questions — how many times. AI assistants do not answer the same question the same way twice, so asking once tells you what one answer looked like at one moment. If a report doesn't say, it was almost certainly asked once.
Ours: stated in the Measurement Passport, on its own line, with a plain-English explanation of why it is the number to look at first. Our baseline is three times per question per engine, and five on the handful of questions a purchase actually follows.
A single percentage from a system that answers differently every time is one moment dressed up as a standing fact. A range tells you what tends to happen, which is the only thing you can plan around.
Ours: every rate is reported as a range. And where an audit was run only once, we say so and report named or not named instead — because a percentage calculated from a single pass would look precise and mean almost nothing.
Every method has limits. A report with no limitations section either didn't look for them or decided not to mention them.
Ours: a standing limitations section in every copy, written into the report generator so it cannot be left out. It says plainly that we do not measure paid placement inside AI answers, and that we measure single questions rather than conversations.
Ask to see one transcript — the actual text an engine returned. If a provider only has scores, scores are all they ever had.
Ours: every finding traces to a captured answer. The probe log, the method appendix and the evidence ledger are all in the report you receive, not held back.
We publish these because we would rather be measured against them than not. If another provider answers all four clearly, that is a good sign about them — the point is that you get to ask.
Our full method is public and ungated — read exactly how the audit is run.
This is the failure we see most often, and it's the reason we don't hand anyone a single grade. A business can look healthy across its category and be completely absent from the handful of questions that actually precede a purchase.
A specialty medical practice we audited was recommended consistently, on every engine, for the narrow specialty it genuinely leads. On a category-wide visibility score, that reads as a win.
On its two biggest money questions — a named elective procedure, and “top-rated” for the specialty — it was named by none of the four engines. Competitors owned both answers outright.
Most AI visibility tools run a broad list of questions and count how often your name comes back. Every question weighs the same, so a large number of easy questions can drown out a small number of expensive ones. The score goes up. The pipeline doesn't.
We ask the questions your buyers actually ask, in three separate layers, and we report each layer on its own:
For local businesses we run each layer at more than one geographic radius as well, because “best X in [city]” and “best X in [suburb]” routinely return completely different companies. Reporting one number across all of it would hide exactly the finding you're paying us to surface.
The practice above is described by sector only. We don't publish client names or findings without written permission — an audit is a catalog of where a business is weakest in public, and that isn't ours to hand to their competitors. The same protection applies to you.
It seemed only fair. We put regulus.digital and our sister company elderhonor.com through the same lenses we sell — and found real problems: a sitemap declared but not serving, an llms.txt that never deployed, a homepage heading that said “Home,” and crawl directives nobody had touched in years.
We fixed all of it the same week, and re-ran the independent checks. It is a strange thing to advertise, but it is the honest version: the audit finds things, including on the people selling it.
See the method we usedCheck any of it yourself — that’s the point of listing it.
The scan and audit show where AI misses you. The AI Content Roadmap is the sequenced, done-with-you plan that fixes it — built from the engines' own follow-up questions, with projected impact we re-measure and prove.
See the AI Content RoadmapIncluded with a Margin Scan, or $1,500 on its own. Either way, you'll know where you stand and how to climb.