We audit what ChatGPT, Claude, Gemini and Perplexity actually say about your business today, rebuild the record they read from, publish it where machines look — and prove it moved in 30 days.
Australian businesses. Licences checked against ABR, NSW Fair Trading and QBCC.
If we sent you a readiness report, that was the free scan — the same rubric, run on your live site, with the real answers assistants gave when we asked them a customer's question. The audit is what happens next: we fix every finding in it, publish the result, and re-measure.
Every line below is a real artifact you can open, check and keep — whether or not you stay with us afterwards.
Your live site scored against the complete rubric — structured data, machine-readable summaries, sitemap and crawler access, server-rendered content, and how complete your contact, hours, services and pricing actually are. Checks that don't apply to your kind of business are excluded rather than scored against you.
We ask each major assistant the questions your customers actually ask — "best {your trade} near {your suburb}", "who does {your service} in {your city}" — and record what came back. Who gets named, who gets cited, and where you appear if you appear at all.
We extract your entire business into one structured record — services, pricing, hours, service areas, credentials, policies, FAQs, response times, lead times, the lot. This is the thing assistants read. Most businesses have never had one, which is exactly why AI improvises about them.
Your ABN and trade licences checked against the actual registers — the Australian Business Register, NSW Fair Trading, QBCC — so the record carries facts rather than assertions. This is the difference between a listing and a reference an assistant will repeat.
The record goes live as schema.org structured data, a machine-readable summary file, and a live agent endpoint that assistants can call directly to get your hours, services, availability and prices. Then we fetch every one of those from the outside and confirm it actually resolves — a published artifact that 404s is worse than none.
We re-run the identical scan and the identical questions 30 days later and hand you a before/after. Same rubric, same prompts, same assistants. You see exactly what moved and what didn't — including the things we couldn't shift, which we'll tell you about plainly.
The only thing we need from you is one conversation and access to confirm what we can't verify ourselves.
Full scan, live visibility test across every assistant, and the baseline we'll measure against.
We walk you through the findings and confirm the facts only you can confirm — prices, policies, what you actually do.
Record built, credentials checked against the registers, everything published and fetched back to prove it's live.
Identical scan re-run, before/after report, and everything handed over in a form you own.
You are not signing up to software. You are buying a piece of work with a defined end.
Prices in AUD, excluding GST. After the audit, ongoing monitoring is optional — monthly re-scans, accuracy drift alerts and a live visibility trend. Cancel any time. See plans →
If your readiness score hasn't measurably improved at the day-30 recheck, you don't pay the second half. Not "we'll make it right" — you keep the work, the record, the published artifacts and the report, and we invoice you nothing further.
What we will not promise is a number of leads. Nobody controls what an assistant says, and anyone who tells you otherwise is selling you something. What we control is whether the machines can read you accurately — and that we guarantee.
No, and the difference is mechanical. SEO competes for a position in a list of links a person then clicks. An assistant doesn't return a list — it returns one answer, composed from whatever it can read and verify about you. Ranking fourth is worthless when there is no fourth. The work is different too: no backlinks, no keyword density. It's structured data, machine-readable facts and a callable endpoint.
Because they're separate pipes. Ranking on Google means Google's crawler understood your pages well enough to place them. It says nothing about whether an assistant can extract your Saturday hours, your callout fee or your licence number and state them confidently to a customer. We regularly scan businesses that sit at the top of Google and are described inaccurately, or not at all, by every major assistant.
Then the audit is more useful, not less. The record is the deliverable, not the website — and it publishes to structured data, a machine-readable summary and a live agent endpoint whether or not you have a site. If you don't have one, we can generate one from the same record at no extra cost, because it's the same information rendered differently.
You do. The record, the structured data and the report are exportable and yours whether you stay with us or not. The monitoring subscription is optional and buys you re-scanning and alerts, not custody of your own information. If you leave, you leave with everything.
Because the checks are external and repeatable. The rubric is published at /assess/methodology, the visibility test uses the assistants' own public interfaces with the questions recorded verbatim, and every published artifact is verified by fetching it from outside our network. You can re-run the free scan yourself, any time, and compare.
It hasn't happened yet, but the honest answer is that we'd tell you on day two and refund the deposit rather than bill you for a fix you don't need. A scan that comes back clean is a good outcome; charging you $2,400 to confirm it is not.
The scan is free and takes about two minutes. If it comes back clean, you've lost nothing. If it doesn't, you'll want to know.