Skip to content

Ten blue links became
 three names. Be one.

A growing share of buyers asks an assistant instead of scrolling results, and it answers with a short list of names. Page-one ranking does not put you on that list.

  • Run the free visibility check
  • See who the assistants name today
  • Get the fix list, in order

Book time with our team

Tell us who you are and we'll bring the right numbers to the call.

Team size

Trusted by teams like yours

What it costs you now

A buyer asks an assistant who to call, and your name is not in the answer.
You rank on page one, and the answer sitting above it never mentions you.
A model describes your company from sources nobody on your side has read.
Nobody on your team owns what the assistants say about you this month.
What we go after

Three things have to be true
before a model says your name.

#1

Named

Be one of the names it says. The list an assistant reads out is short and it is not an auction. Getting onto it is a content and corroboration problem, not a bidding one.

#2

Described right

And described the way you sell. Trades, service area, licence, hours. A model repeats whatever its sources agree on, and out in the wild they rarely agree.

#3

Cited

Linked, not just mentioned. AI Overviews and Perplexity attach their sources. A page written so a paragraph can be lifted out of it is the page they attach.

What we watch

We could argue about AI search.
Easier to go and look.

The four assistants your buyers ask, the prompts they actually type, and what the first look costs you.

  • 4 Assistants checked: ChatGPT, AI Overviews, Perplexity, Gemini
  • 30+ Prompts per market, re-asked every month
  • Free The first visibility check, transcripts included
Why now

The ground is open, and it does not stay open.

Every competitor in your market bids on the same keywords. Almost none of them have checked what an assistant says when a buyer asks who to call. That ground closed in classic search around 2012; in the assistants it is still wide open.

  • You are in the answer or you are invisible. There is no page two.
  • Being described correctly counts for as much as being there at all.
  • The work compounds. Every corroborating source keeps paying.
The four we track

  • ChatGPT

    The first stop for a growing share of buyer research.

  • AI Overviews

    Sits above the organic results you already fight for.

  • Perplexity

    Citation heavy, and the easiest of the four to earn a place in.

  • Gemini

    Pulls from your Google Business data, so listing accuracy shows up directly.

The offer

The first look costs nothing.

We put your trade and your metro to all four assistants, the way a buyer would put it, and send you back what came out.

  • The transcripts, unedited. You read what was said, not a summary of it.
  • Where you stand in each one. Named, missing, or described wrong.
  • The fix list. What we would change, in the order we would change it.
The work

Four layers, run in order.

  1. The classic half

    Service-area pages, schema, reviews and citations: unglamorous, still decisive.

  2. Retrievability

    We rewrite pages so a model can lift a paragraph out of one and be right.

  3. Corroboration

    Directories, licence records and review sites made to agree on your name and numbers.

  4. Monitoring

    The four assistants re-asked every month, on the prompts your buyers type.

Which layer leads on your account is decided by the check, not by a template.

Questions
Is this just SEO with a new name?

It shares the classic half. But ranking means being one of ten links someone scrolls, and being cited means being one of three names a model speaks. Different content requirements, and far less competition.

How long before it moves?

Retrievability work often shows inside a month. Corroboration is slower. Six months is the honest horizon for a cold market, and we would rather say that than sell you a date.

What if an assistant says something wrong about us?

More common than being absent, and more damaging. Fixing it means correcting the sources the model pulls from, which is a large part of this work.

Which prompts do you use?

The ones your buyers type, plus the questions your own sales calls keep repeating. Thirty-plus per market, agreed with you and re-asked every month so the answers are comparable.

Is there a contract?

Month to month. The pages, the schema and the corrected listings are yours, and they keep working whether we are still on the account or not.

See what they say about you
before you fix a thing.

We run your trade and your metro through all four assistants, send you the transcripts, and tell you what we would change first. Nothing to install, and no charge for the look.

Get the free visibility check See the four layers