Search & AI Visibility · Generative Engine Optimization (GEO)

Become a source a model is willing to name

Language models do not “visit” your brand. They retrieve passages that look like answers, from entities they trust. This engagement is the unglamorous citation work: a real author, first-hand proof, consistent facts across the web, and pages written the way a person asks out loud.

  • Author is a named person with a role, not a brand ghost
  • Claims that cannot be sourced are cut, not softened into marketing
  • Organization name, URL, and profiles are consistent so the model does not split you
  • We probe branded and non-branded prompts and write down who got cited

What you get

Deliverables, not slideware

  • 01

    Entity and fact lock

    One Organization, one canonical URL, the same short description on the site, LinkedIn, and anywhere else you already exist. Conflicting bios are treated as bugs.

  • 02

    Evidence pass

    Each priority page gets at least one first-hand proof: a named client where we have permission, a method we actually run, or a number with a source. No “industry-leading” filler.

  • 03

    Citation-ready passages

    Short, conversational answers to long-tail questions, with the kind of specific that is hard to fabricate. Original data only when we collected it.

  • 04

    Visibility probe

    A written set of branded and non-branded prompts, run in the tools your buyers use, with who was cited. This is a snapshot, not a ranking product. We will not invent a citation percentage.

geo-probe.md
  • 01Prompt: “Who builds custom operating systems for owner-led companies in Raleigh?”
  • 02Cited: (the domain that appeared)
  • 03Prompt: “What is generative engine optimization for a B2B service company?”
  • 04Cited: (the domain that appeared)
  • 05Note: snapshot on the date of the diagnostic, not a live score
Sample of the probe sheet. A dated snapshot, not a live score.

How it runs

What actually happens, step by step

No discovery theatre. Each step ends with something you can read or use.

  1. 01

    Lock the entity

    Name, URL, author, sameAs profiles. Kill the second About paragraph that disagrees with the first.

  2. 02

    Cut unsourced claims

    If a number cannot be opened in a tab, it leaves the page.

  3. 03

    Rewrite for speech

    Long-tail, question-shaped, specific. No keyword stuffing. No “revolutionary platform.”

  4. 04

    Probe and record

    Run the prompt set. Record who was named. Use that list to choose the next page, not to print a trophy.

Tangible

What lands in your hands

Named objects with a format and a week attached, so the handover is easy to picture.

  • Visibility audit

    Where you appear in search and in assistant answers, and where you do not.

    PDF · 11 pagesWeek 2
  • Answer page plan

    The questions worth owning, ranked by intent and difficulty.

    Prioritised brief setWeek 2
  • Tracking dashboard

    Rankings, citations and assisted conversions in one place, refreshed weekly.

    Live URLWeek 3
Sample

Search & AI Visibility Audit

Where you appear, where you do not, and the pages that would move it.

11 pp
Download the sample (PDF)

This is the real template, with client figures replaced by representative ones. Read it before you talk to us — if the format is not useful to you, the engagement will not be either.

Before and after

What the change looks like in the week

Red is the cost you carry today. Green is what the system gives back.

Today

  • Three different one-line descriptions of the company exist on the web
  • Pages are polished and empty of a person who did the work
  • Stats have no link
  • Nobody has asked ChatGPT the questions your buyer asks

After

  • One entity, one set of facts
  • A named author and a piece of first-hand work on the page
  • Every hard number has a source or is gone
  • You have a dated list of prompts and who got the citation

Objections

Answered before you ask

Can you get us cited by ChatGPT?

We can make you citable. We cannot order a model to name you. Anyone selling a guaranteed ChatGPT citation is selling a thing they do not control.

Is this digital PR?

Partly. Mentions on sites a model already trusts help. We will say when a guest post or a named case is the next move, and when another page on your own site is enough.

Do we need original research?

It helps, and it is not required on day one. First-hand process, named work, and clean facts already beat a generic roundup.

One next step, and it is a paid one on purpose

The AI Opportunity Diagnostic is a fixed-scope engagement. You leave with a ranked plan you can act on with us or without us.

Still sizing the problem? Write to us in your own words first. A person reads it.