Can you guarantee ChatGPT will recommend us?

No, and neither can anyone else. The engines re-rank without notice and nobody outside the labs controls the output. What we guarantee is the work in the scope, a measured baseline, and a monthly report that shows a flat month as flat.

Is the technical work really finished after two cycles a quarter?

Two cycles a quarter is what a site this size needs once the foundation is in. The audit still refreshes monthly, so anything that breaks or regresses gets caught inside four weeks and goes into the next cycle. What we will not do is bill you monthly for a technical retainer that has nothing left to fix.

We already have an SEO agency. Does this conflict?

Usually not. Classic SEO optimises for a ranked page of links; this optimises for being the source a model quotes when there is no page of links. There is overlap in the technical layer, so we ask for read access to what they have shipped and stay out of their way.

Is llms.txt part of this?

It ships in week one because it costs us an hour, and then we stop talking about it. There is no public evidence that any major engine reads it as a ranking input. If an agency has built its pitch on llms.txt, schema-for-AI and content chunking, you are being sold plumbing as strategy.

Who actually does the work?

The people on the call. Placements and Reddit participation are done under real names by people who already have standing in those communities, which is the whole reason they land. Nothing is offshored to a content farm.

What happens at the end of the 90 days?

You keep everything: the assets, the entity work, the placements and the baseline data. You can stop, renew at the same scope, or move to Expansion. There is no auto-renewal and no clause that makes leaving expensive.

Tier 02of 03

Your competitors arealready the answer.This is the 90 days you take it back.

Ten cite-ready assets and up to five listicle placements a month, measured against the same hundred prompts.

Own the answers buyers get.

Domination

MOATWORKS

Scaling B2B SaaS, roughly $5M–$10M ARR, putting real budget behind search for the first time.

$2,999/month

90-day termAudit included, not billed on top

See how this compares↓

Start Dominating⟶

Included

Every line, itemised.

$2,999 a month

no mystery

just the raw scope

$2,999 a month

no mystery

just the raw

scope

Scope locked at the end of week two. The number you signed is the number you pay.

What you hold at day 90

15placements live
Listicle and roundup placements on the sources the five engines demonstrably cite, published under real bylines.
30cite-ready assets
Each one written against a specific prompt your baseline audit showed you losing, dated and sourced well enough to quote.
30 – 40queries you now win
Tracked prompts where an engine names you and did not at baseline, measured monthly against the same fixed set.
1board-ready number
AI sessions through to demos and signups in a single monthly report, so the spend has a line you can defend.

The full scope

Every line we are on the hook for, with the quantity attached. Anything not on this list is not in the price, and we will say so before you sign.

Strategy

AI visibility audit: 100 prompts across 5 engines, refreshed monthly
Buyer query research: 100 queries, top 40 prioritised
Competitor citation analysis: At onboarding, then monthly
Competitor displacement tracking: Monthly: who is cited where, and when that shifts
Keyword and content gap research: Every month, against the tracked prompt set
GA4 AI session tracking: AI sessions segmented, goals and conversions configured

Content

Cite-ready assets: 10 per month, 30 across the term
Comparison and alternatives pages: Included in the monthly 10, prioritised by lost prompts
Existing page revamps: Up to 4 per month where a page already ranks but is never quoted
FAQ and answer formatting: Applied across every asset we publish or revamp

Technical SEO

Technical audit: At onboarding, then refreshed monthly
Technical fix cycles: 2 per quarter, scoped from the audit and shipped by us
Schema and structured data: Up to 15 pages per quarter
Internal linking: Up to 90 new links per quarter, mapped and implemented in your CMS
Entity and knowledge graph: Build at onboarding, then ongoing maintenance
AI entity snippets: Included, reviewed each month

Off-page

Listicle and roundup placements: Up to 5 per month, 15 across the term
Guest posts / Digital PR: 2 per month, placed under real bylines
Review platform presence: G2, Capterra and category pages kept current and correctly categorised
Reddit community seeding: 4 posts and 45–60 comments per month, written by us, posted under real accounts
Placement budget: Included in the $2,999, nothing billed separately

Reporting

AI queries tracked: 100 prompts, re-run monthly against a fixed baseline
Competitor share of voice: Monthly, against a named competitor set
Monthly performance report: GSC, GA4, citations and rankings in one view
Pipeline attribution: AI sessions through to demos and signups, board-ready
Placement report: Every placement listed with live URL, domain and anchor text

Working together

Calls: 2 per month, plus the 90-day scoping call at the start
Scope lock: Fixed at the end of week two, no change orders after that

The 90 days, in order

Nothing is published before the baseline exists. Placements are the slow part, so they start early and finish last.

.01Days 1–14
Baseline and foundation
Nothing gets published before we know where you stand. The audit runs, the first technical cycle ships, and scope is locked against what the audit found.
- Audit runs across 100 prompts and five engines
- Per-engine playbook and competitor gap analysis delivered
- First technical fix cycle and schema pass shipped
- 90-day scope locked, no change orders after this point
.02Days 15–45
Entity work and first assets
You become a resolvable entity with facts that agree with each other, and the first assets go live against the prompts you were losing.
- Entity map, knowledge graph and Wikidata entry built
- Hallucination fixes filed at the source, per engine
- First 10 cite-ready assets published
- First 5 listicle placements live, Reddit seeding running
.03Days 46–90
Placements compound
Citations take time to be crawled, trusted and repeated. This is the stretch where the earlier work starts showing up inside answers.
- Placements reach 15 across the term
- 20 further cite-ready assets published
- Second technical fix cycle shipped
- Attribution report ties AI-sourced sessions to pipeline

The work

Five surfaces, explained.

An answer engine

is not a search engine

five things decide

whether it names you

An answer

engine

is not a

engine

five things

decide

whether it

names you

Domination covers all five. Most of the market sells you the cheap one.

.01

Technical SEO

Answer engines can only cite what they can parse. This is the plumbing: how your pages render, how they are structured, and how a claim is packaged so a model can lift it whole and attribute it to you. It is the floor of the work rather than the strategy, which is why it runs on a fixed cycle instead of being re-sold every month.

In Domination

Two fix cycles a quarter, schema on up to 15 pages, and up to 90 new internal links mapped and implemented in your CMS.

Technical Audit: At onboarding, then refreshed every month against what changed.
Fix Cycles: Two per quarter, scoped from the audit and shipped by us.
Schema & Structured Data: Up to 15 pages a quarter, so engines parse rather than guess.
Internal Linking: Up to 90 new links a quarter, mapped and implemented in your CMS.
Crawler Access: GPTBot, ClaudeBot and PerplexityBot allowed, then verified in your logs.
Server Rendering: Content present in the HTML, not assembled after JavaScript runs.

Entity and authority: making a company resolvable and consistent across the sources models trust

.02

Entity & Authority

Models resolve companies as entities, not as URLs. If your business is not a resolvable entity with consistent facts across the sources models trust, you get merged with a competitor, described with last year's positioning, or invented outright.

In Domination

Entity map, knowledge graph and Wikidata work at onboarding, then maintained monthly, plus fixes for every hallucination the audit surfaces across the five engines.

Entity Mapping: Company, products and people defined as things a model can resolve.
Knowledge Graph: The same facts wherever an engine goes looking for them.
Wikidata Entry: A structured, citable record the models already trust.
AI Entity Snippets: The short, quotable definition of you that engines reach for first.
Source Hardening: Correcting the outdated third-party pages engines keep quoting.
Hallucination Fixes: Find what engines get wrong about you, then fix it at the source.

Earned media and off-page: the third-party sources answer engines pull from when recommending a vendor

.03

Earned Media & Off-page

This is the lever. Models cite third-party sources far more often than they cite your own site, so being recommended means being present in the roundups, listicles and threads the engines pull from. It is the slowest part of the work and the part that compounds.

In Domination

Up to 5 listicle and roundup placements a month, which is 15 across the term, two guest posts or digital PR placements a month, plus four Reddit posts and 45 to 60 comments a month, written by people who already post there.

Listicle Placement: The "best tools for X" pages models compress straight into answers.
Roundup & Review Sites: G2, Capterra and the category pages engines summarise.
Guest Posts & PR: Two placements a month, secured under real bylines.
Reddit Seeding: Four posts and 45 to 60 comments a month, in the threads your buyers read.
Wikipedia Page: Only where notability is genuine. We will tell you when it is not.

Content and answers: prompt-led assets written to be quoted verbatim by answer engines

.04

Content & Answers

Not blog volume. Prompt-led assets built from the queries that failed in your audit: comparison pages, alternatives pages and data-dense answers written to be quoted verbatim, with the losses admitted rather than hidden.

In Domination

10 cite-ready assets a month, which is 30 across the term, each mapped to a prompt the audit showed you losing.

Comparison Pages: You against the alternatives, including where you lose.
"vs" Pages: The exact query shape buyers type right before they shortlist.
Prompt-Led Content: Written against prompts we watched you fail in the audit.
Alternatives Pages: The page that catches everyone leaving a competitor.
FAQ Optimization: The answer in the first sentence, not the fifth paragraph.
Cite-Ready Format: Dated, sourced and specific enough for a model to quote.

Tracking and proof: monthly prompt re-runs across five engines with source and pipeline attribution

.05

Tracking & Proof

The same 100 prompts run against five engines every month, so movement is measured rather than claimed. You see the score, the share of answers, the exact URLs the engines cited, and what any of it did to pipeline.

In Domination

A monthly report with the full prompt set re-run, competitor movement, every placement listed with its live URL, and AI-sourced pipeline attribution.

Moat Score: One number for how often engines name you unprompted.
Share of Voice: Your share of answers against a named competitor set.
Source Tracking: The specific URLs an engine cited to build its answer.
Pipeline Attribution: AI sessions through to demos and signups, not impressions.
Competitor Watch: An alert when a rival starts winning answers you held.
Placement Report: Every placement with its live URL, domain and anchor text.

Comparison

Same category. Different work.

Every agency says

the same five words

compare the deliverables

Every agency

says

the same five

words

compare the

deliverables

Where we do less than the market, it says so.

The Domination plan compared with a typical GEO agency retainer, grouped by service category.
Capability	MoatWorks Domination $2,999/mo · 90-day term	Typical GEO agency $2,500–4,000/mo · 12-month lock
01Technical SEO
Technical fix cycles	DominationIncluded2 a quarter, scoped from the audit and shipped by us	Typical agencyPartly, or not as soldRe-sold every month as the ongoing deliverable
Internal linking	DominationIncludedUp to 90 links a quarter, implemented in your CMS	Typical agencyPartly, or not as soldA spreadsheet of suggestions handed to your team
llms.txt and schema-for-AI	DominationPartly, or not as soldShipped because it is cheap, and we say plainly it is not the lever	Typical agencyPartly, or not as soldPositioned as the core of the offer, with no evidence behind it
Content chunking	DominationNot includedNot a service. No published evidence it changes retrieval	Typical agencyPartly, or not as soldSold as a line item, priced as strategy
02Entity & Authority
Entity and knowledge graph	DominationIncludedBuilt at onboarding, then maintained monthly	Typical agencyNot includedRarely touched, it sits outside a content retainer
Hallucination correction	DominationIncludedTracked per engine monthly, fixed at the third-party source	Typical agencyNot includedNot measured, so never fixed
03Earned Media & Off-page
Listicle and roundup placements	DominationIncludedUp to 5 a month, 15 across the term	Typical agencyPartly, or not as soldGuest posts on domains bought for DA, which models do not read
Reddit and communities	DominationIncluded4 posts and 45–60 comments a month, by people who already post there	Typical agencyNot includedAvoided as unscalable, or automated until the account is banned
Review platforms	DominationIncludedG2 and Capterra profiles kept current and correctly categorised, because engines quote them	Typical agencyNot includedOut of scope, or handed back to your own marketing team
04Content & Answers
Volume and intent	DominationIncluded10 assets a month, each mapped to a prompt you were losing	Typical agencyPartly, or not as sold20+ posts a month, chosen by keyword volume to fill a quota
Format	DominationIncludedDated, sourced, data-dense, quotable in a single passage	Typical agencyNot includedModel-written, unsourced, too generic to be worth citing
Comparison pages	DominationIncludedWritten with the losses admitted, which is why they get quoted	Typical agencyPartly, or not as soldWritten as sales copy, which is why they get ignored
05Tracking & Proof
Measurement	DominationIncluded100 prompts across five engines, re-run monthly	Typical agencyNot includedScreenshots of one favourable ChatGPT answer
Pipeline attribution	DominationIncludedAI sessions through to demos and signups, monthly	Typical agencyNot includedTraffic, impressions and a rising-line chart
Commitment	DominationIncluded90 days, scope locked before anything moves	Typical agencyNot included12-month contract, scope decided as it goes

IncludedIncluded at this price
Partly, or not as soldPresent, but not as it is sold
Not includedNot included

“Typical GEO agency” describes the pattern we see across public pricing pages, pitch decks clients forward us, and the audits we run on their work, not any single company. If yours does better than this column, keep them.

Four reasonsnot to buy this.

A $9K commitment sold to the wrong company churns in month two and costs us the case study. It is cheaper for both of us to disqualify you now.

All three tiers

01
You don't know where you stand yet.
If nobody has measured how the engines currently answer for your category, a retainer is premature. Buy the audit, read it, then decide. The $800 comes off if you upgrade.
Foundation, $800 one-time⟶
02
AI visibility is already a board metric.
Past roughly $20M ARR, with a number being reported upward every quarter, the placement and asset volume here is too thin to move a market you are already competing hard in.
Expansion, $4,999 a month⟶
03
You want a guarantee of being cited.
Nobody can sell you that, and anyone who does is guessing. What is guaranteed is the work in the scope, a measured baseline, and a monthly report honest enough to show a flat month as flat.
04
You want 20 blog posts a month.
That is a content shop, and there are cheaper ones. This is 10 assets built against failed prompts, plus the off-page work that makes them worth citing.

Objections

Answered straight.

The questions we get on the second call.

Because citations do not compound in 30. Placements have to be published, crawled, and then repeated by a model before they show up in an answer, which is a 6 to 10 week arc on a good run. A monthly rolling deal would let us bill you through the part where nothing is visible yet, and let you cancel right before the part where it is.

Ninety days from now, someone is the answer.

$2,999 a month, 90-day term, audit included and scope locked before anything moves. The first call is a scoping call, not a pitch.

Start Dominating⟶

The Domination plan compared with a typical GEO agency retainer, grouped by service category.
Capability	MoatWorks Domination $2,999/mo · 90-day term	Typical GEO agency $2,500–4,000/mo · 12-month lock
01Technical SEO
Technical fix cycles	DominationIncluded2 a quarter, scoped from the audit and shipped by us	Typical agencyPartly, or not as soldRe-sold every month as the ongoing deliverable
Internal linking	DominationIncludedUp to 90 links a quarter, implemented in your CMS	Typical agencyPartly, or not as soldA spreadsheet of suggestions handed to your team
llms.txt and schema-for-AI	DominationPartly, or not as soldShipped because it is cheap, and we say plainly it is not the lever	Typical agencyPartly, or not as soldPositioned as the core of the offer, with no evidence behind it
Content chunking	DominationNot includedNot a service. No published evidence it changes retrieval	Typical agencyPartly, or not as soldSold as a line item, priced as strategy
02Entity & Authority
Entity and knowledge graph	DominationIncludedBuilt at onboarding, then maintained monthly	Typical agencyNot includedRarely touched, it sits outside a content retainer
Hallucination correction	DominationIncludedTracked per engine monthly, fixed at the third-party source	Typical agencyNot includedNot measured, so never fixed
03Earned Media & Off-page
Listicle and roundup placements	DominationIncludedUp to 5 a month, 15 across the term	Typical agencyPartly, or not as soldGuest posts on domains bought for DA, which models do not read
Reddit and communities	DominationIncluded4 posts and 45–60 comments a month, by people who already post there	Typical agencyNot includedAvoided as unscalable, or automated until the account is banned
Review platforms	DominationIncludedG2 and Capterra profiles kept current and correctly categorised, because engines quote them	Typical agencyNot includedOut of scope, or handed back to your own marketing team
04Content & Answers
Volume and intent	DominationIncluded10 assets a month, each mapped to a prompt you were losing	Typical agencyPartly, or not as sold20+ posts a month, chosen by keyword volume to fill a quota
Format	DominationIncludedDated, sourced, data-dense, quotable in a single passage	Typical agencyNot includedModel-written, unsourced, too generic to be worth citing
Comparison pages	DominationIncludedWritten with the losses admitted, which is why they get quoted	Typical agencyPartly, or not as soldWritten as sales copy, which is why they get ignored
05Tracking & Proof
Measurement	DominationIncluded100 prompts across five engines, re-run monthly	Typical agencyNot includedScreenshots of one favourable ChatGPT answer
Pipeline attribution	DominationIncludedAI sessions through to demos and signups, monthly	Typical agencyNot includedTraffic, impressions and a rising-line chart
Commitment	DominationIncluded90 days, scope locked before anything moves	Typical agencyNot included12-month contract, scope decided as it goes

01Technical SEO

Technical fix cycles

Domination2 a quarter, scoped from the audit and shipped by us

Typical agencyRe-sold every month as the ongoing deliverable

Internal linking

DominationUp to 90 links a quarter, implemented in your CMS

Typical agencyA spreadsheet of suggestions handed to your team

llms.txt and schema-for-AI

DominationShipped because it is cheap, and we say plainly it is not the lever

Typical agencyPositioned as the core of the offer, with no evidence behind it

Content chunking

DominationNot a service. No published evidence it changes retrieval

Typical agencySold as a line item, priced as strategy

02Entity & Authority

Entity and knowledge graph

DominationBuilt at onboarding, then maintained monthly

Typical agencyRarely touched, it sits outside a content retainer

Hallucination correction

DominationTracked per engine monthly, fixed at the third-party source

Typical agencyNot measured, so never fixed

03Earned Media & Off-page

Listicle and roundup placements

DominationUp to 5 a month, 15 across the term

Typical agencyGuest posts on domains bought for DA, which models do not read

Reddit and communities

Domination4 posts and 45–60 comments a month, by people who already post there

Typical agencyAvoided as unscalable, or automated until the account is banned

Review platforms

DominationG2 and Capterra profiles kept current and correctly categorised, because engines quote them

Typical agencyOut of scope, or handed back to your own marketing team

04Content & Answers

Volume and intent

Domination10 assets a month, each mapped to a prompt you were losing

Typical agency20+ posts a month, chosen by keyword volume to fill a quota

Format

DominationDated, sourced, data-dense, quotable in a single passage

Typical agencyModel-written, unsourced, too generic to be worth citing

Comparison pages

DominationWritten with the losses admitted, which is why they get quoted

Typical agencyWritten as sales copy, which is why they get ignored

05Tracking & Proof

Measurement

Domination100 prompts across five engines, re-run monthly

Typical agencyScreenshots of one favourable ChatGPT answer

Pipeline attribution

DominationAI sessions through to demos and signups, monthly

Typical agencyTraffic, impressions and a rising-line chart

Commitment

Domination90 days, scope locked before anything moves

Typical agency12-month contract, scope decided as it goes

The questions we get on the second call.