What a GEO Consultant Actually Does

Published August 2026.

TL;DR: “GEO consultant” is a title anyone can print on a business card right now — there’s no license, no shared curriculum, and no agreed job description. The actual work, when it’s real, is three things: an entity and structured-data audit, a citation-tracking baseline across ChatGPT/Perplexity/Claude, and a prioritized fix list ranked by effort versus citation impact. If a proposal skips straight to a monthly retainer with no baseline and no audit, that’s the tell it’s not GEO work — it’s an SEO retainer with a new label on it.

[Operator’s read] I run GEO audits for operators who don’t have a marketing team to hand this to, alongside this site, a productized-service business, and a course. This isn’t a survey of the category — it’s a description of what I actually deliver, and what I’d tell a friend to check before hiring anyone for it, including me.

---

Table of contents

Open Table of contents

• Why the title is confusing right now
• The three things real GEO work actually delivers
  • 1. An entity and structured-data audit
  • 2. A citation baseline, measured before any work starts
  • 3. A prioritized fix list, ranked by effort versus citation impact
• How this differs from an SEO retainer with a new label
• Red flags to check before hiring anyone for this
• Who actually needs this versus who can skip it
• FAQ
  • Is GEO consulting the same thing as an “AI citation service”?
  • How long does a real GEO audit take?
  • Can I do this myself instead of hiring someone?
  • What should the first deliverable from a GEO consultant look like?
• The operator’s bottom line

Why the title is confusing right now

“SEO consultant” has meant roughly the same thing for two decades: rankings, backlinks, technical audits, content strategy, all measured against Google’s organic results. Anyone hiring one has a rough mental model of the deliverable before the first call.

“GEO consultant” doesn’t have that yet. The discipline — getting cited inside ChatGPT, Perplexity, Google’s AI Overviews, and Claude’s web search — is maybe two years old as a paid service category. That means two things are true at once: the skill is genuinely new and valuable, and the label is loose enough that an SEO agency can relabel its existing retainer “GEO” without changing the work inside it.

The way to tell the difference isn’t the pitch deck. It’s whether the engagement produces something SEO work doesn’t: a citation baseline, an entity graph audit, and structured data that’s actually verified against what the page says — not just present.

---

The three things real GEO work actually delivers

1. An entity and structured-data audit

AI engines don’t rank pages the way a search index does — they build an entity graph and decide who’s a credible source about a topic. That graph is only as clean as your structured data. A real audit checks:

• Whether Person or Organization schema exists on the pages that need to anchor citations, with sameAs links that actually resolve to real, consistent profiles
• Whether the schema on a page matches the visible content — a FAQPage block that doesn’t match the FAQ text on the page is a trust signal working against you, not for you
• Which schema types are missing entirely on pages that should carry them (see schema markup for AI engines: the types that punch above their weight for which ones actually move citations versus which ones are decorative)

This is mechanical, verifiable work. If a “GEO audit” doesn’t include a line-by-line pass of your actual JSON-LD against your actual page content, it wasn’t an audit — it was a keyword list with a new name.

2. A citation baseline, measured before any work starts

You can’t show impact without a starting point. A real GEO engagement runs a fixed set of prompts — “best [category] for [your ICP],” direct comparison queries, “who does X” — across ChatGPT, Perplexity, and Claude, and logs whether you’re cited, what’s said about you when you are, and who’s cited instead when you’re not. That log is the baseline everything else gets measured against.

This is the same check I run on a recurring cadence for my own properties, described in GEO for solo operators — the only difference in a client engagement is that the first run happens before any fixes ship, specifically so there’s something to compare against later.

3. A prioritized fix list, ranked by effort versus citation impact

The output isn’t a 40-page report nobody reads. It’s a short, ordered list: which structural fixes are one-time and high-leverage (entity schema, a direct-answer block at the top of key pages, FAQ schema that matches real content), which are recurring and should be handed to an agent rather than a person, and which tactics — buying links, chasing every low-authority directory, paying a “citation service” — don’t belong on the list at all because there’s no verified mechanism by which they move a model’s output.

That last category matters more than it sounds like it should. Part of the job is telling a client what not to spend money on.

---

How this differs from an SEO retainer with a new label

The overlap is real — clean structured data, a fast site, and genuinely useful content help both disciplines, and a good consultant won’t pretend otherwise. The differences are in what gets measured and how the engagement is structured:

	SEO consulting	Real GEO consulting
Baseline	Keyword rank tracking	Citation log across ChatGPT, Perplexity, Claude
Primary audit	Backlinks, technical crawl, on-page	Entity graph, schema-to-content match
Success metric	SERP position, organic clicks	Cited or not, and what the model said
Recurring cadence	Monthly content calendar	Weekly citation spot-check, staleness detection
Deliverable shape	Ranking report	Fix list ranked by citation impact

If a proposal’s “GEO deliverables” section is a keyword-density checklist with “AI search” pasted into the header, it’s the left column wearing the right column’s name.

---

Red flags to check before hiring anyone for this

• No baseline citation run before the engagement starts. Without one, no report six months from now can honestly claim credit for anything.
• A retainer with no defined audit phase. Structural fixes are mostly one-time work. A proposal that’s monthly-retainer-only, with no distinct “fix these specific things first” phase, is pricing for open-ended hours instead of a defined outcome.
• Any promise to guarantee citations at a given price point. There’s no mechanism by which a paid service makes a model cite a specific domain — the same is true whether it’s called an “AI citation service” or folded into a GEO retainer.
• No mention of structured data verification — only that schema “exists,” not that it matches the page it sits on.
• Reporting that never distinguishes “cited” from “ranked.” Those are different outcomes with different mechanics; a report that conflates them is measuring the wrong thing.

---

Who actually needs this versus who can skip it

A business with a handful of pages, a consistent bio across platforms, and basic schema in place probably doesn’t need a paid engagement — the one-time structural work in GEO for solo operators covers most of the ground in an afternoon.

Where a consultant earns their fee is scale and ambiguity: a site with hundreds of pages and no idea which ones are structurally broken, a company entering a category where competitors are already being cited and you’re not, or a team that needs the citation-tracking cadence built and handed off rather than run ad hoc. The audit-first structure above is what makes the spend defensible either way — you’re paying for a specific, verifiable finding, not a promise.

---

FAQ

Is GEO consulting the same thing as an “AI citation service”?

No, and the distinction matters. A GEO audit identifies structural problems — missing or mismatched schema, weak entity signals, thin direct-answer content — that you or your team can fix. A paid “citation service” claims to get a model to cite you directly for a fee, which has no verified mechanism behind it. A legitimate GEO consultant will tell you the second category doesn’t work, not sell it to you.

How long does a real GEO audit take?

For a site in the tens to low hundreds of pages, a structural audit plus a baseline citation run typically takes one to two weeks — most of that is the citation baseline, which needs to run across multiple engines and multiple prompt variations to be reliable, not a single spot-check.

Can I do this myself instead of hiring someone?

Yes, for a solo operation or small site — see GEO for solo operators for the exact one-time fixes and the agent-run recurring checks that replace a consultant’s ongoing work. The case for hiring out gets stronger as page count and organizational complexity grow, not because the tactics change, but because someone has to do the audit work across hundreds of pages and keep the citation log current.

What should the first deliverable from a GEO consultant look like?

A citation baseline (what’s cited today, across which engines, for which prompts) and a structural audit finding, not a strategy deck. If the first thing you receive is a slide about “the opportunity of AI search” rather than a specific list of what’s broken on your specific pages, ask what was actually audited.

---

The operator’s bottom line

“GEO consultant” will mean something more settled in a few years, the way “SEO consultant” does now. Until then, the way to evaluate one is to ignore the title and check the deliverables: a citation baseline measured before the work starts, an entity and structured-data audit that verifies schema against actual page content, and a fix list ranked by effort versus citation impact — not a retainer sold on the promise of “AI visibility” with nothing to measure it against.

Want a hands-on GEO audit structured this way? See how I run these engagements, or book a 30-minute session to talk through your specific site first.

---

Related: GEO for solo operators: getting cited by AI search · Schema markup for AI engines: the types that punch above their weight · How to get your brand cited in ChatGPT answers