How to Sell AI Visibility Services: Packaging, Pricing, and Positioning #

You sell AI visibility the same way you sell any serious SEO expansion: name the new win condition, prove the gap with a query bank, then productize the fix into audit + build + retainer tiers — not by rebranding last year's link package as "GEO." Agencies that win this work in mid-2026 talk about citations inside ChatGPT, Perplexity, Google AI Overviews, and Google AI Mode the way they used to talk about rankings and CTR. Agencies that lose it either overpromise placements or bury the offer under jargon.

I'm William Spurlock — AI Solutions Architect, Fractional AI CTO, and SEO-certified since 2021. I build AI-visibility-ready sites and help agencies and consultants productize GEO / AEO / AIO work for buyers who still think in blue links. This pillar is written for the agency and consultant buyer register: you already run SEO retainers, you see clients' informational CTR soften, and you need a clear way to package, price, and position the second optimization layer without inventing fake ROI.

This is the pillar for the AI Visibility for Agencies & Consultants cluster. It answers three questions you get every week — how to sell the work, how to explain GEO to an SEO-only client, and how to package and price it — then expands into offer architecture, objection handling, audit deliverables, measurement that survives a procurement call, and FAQ answers you can reuse in proposals. For the keep / drop / add map your account team should internalize first, see AI Visibility vs Traditional SEO. For the content-shape changes you will sell as scope, see GEO vs SEO: what actually changes in how you create content.

---

How do I sell AI visibility services to my clients? #

Sell the gap between "we still rank" and "we never get named," then attach a productized path from audit → extractability fixes → citation retainer — do not open with acronyms. Your client already paid for SEO. They do not need a TED talk on Generative Engine Optimization. They need you to show that buyers asking ChatGPT, Perplexity, or Google AI Mode for a shortlist never hear their brand, even when Search Console looks fine.

The sales conversation that works #

I coach agencies to run a four-beat conversation. Keep it under twenty minutes on the discovery call; put the receipts in a one-pager afterward.

1. Surface the symptom they already feel. Impressions up, clicks down on informational queries. Pipeline from blog softer while brand search holds. Competitors with thinner sites getting named in AI answers.

2. Name the new win condition. Traditional SEO optimizes for rank position and click share. AI visibility optimizes for extraction and citation inside synthesized answers. Both still need crawlable pages and real expertise. Only one cares about being the source the answer names.

3. Prove the gap with a mini query bank. Pick 15–25 buyer questions. Run them in ChatGPT (with web search where available), Perplexity, Google AI Overviews / AI Mode, and note who gets cited. Screenshot the absences. This is the demo — not a deck.

4. Offer a path, not a panic rebuild. Audit (baseline + prioritized fixes) → page and schema sprint (extractability on money URLs) → monthly citation ops (query bank, content, measurement). Price each stage. Never sell "guaranteed AI Overview placements."

What you are actually selling #

Offer layer	Client outcome	Typical deliverables
Diagnosis	Know where they are invisible	Query bank, citation scorecard, technical/entity gaps
Foundation	Money pages become extractable	Answer-first rewrites, FAQ, schema, entity cleanup
Content ops	Steady citation attempts on priority questions	Question-cluster briefs, dual-purpose pages, internal links
Monitoring	Scoreboard for renewals	Biweekly citation checks, SERP/AI surface notes, CRM-assisted pipeline notes

If your proposal only sells "more content," you are selling 2021 SEO with a new sticker. If your proposal only sells a monitoring dashboard, you are selling anxiety. Sell the operating system: questions → extractable pages → corroboration → measurement.

Who buys this fastest #

• B2B services and SaaS where buyers research with AI before booking demos
• Local and regional professionals (legal, medical, home services) losing discovery to answer engines on "best / near me / vs" queries
• Ecommerce and catalog brands that need product entities and merchant data readable by AI surfaces
• Existing SEO retainers that already trust you — expansion is easier than net-new cold outreach

Who buys slowly: founders who only measure organic sessions and refuse a second KPI. Do not force the sale. Educate, leave the query-bank sample, and revisit after the next CTR dip.

Positioning lines you can steal (and adapt) #

Use plain language. Skip hype framing. Specificity closes.

• "We still protect your Google rankings. We also make sure ChatGPT and Perplexity can name you when a buyer asks for a shortlist."
• "SEO gets you into the results. AI visibility gets you into the answer."
• "Your blog still ranks. The Overview answers the question without clicking you. We fix extractability so you become the cited source."
• "This is not a replacement retainer. It is a second optimization layer on the URLs that already make you money."

For a DIY-shaped proof artifact clients can understand before they sign, point them at the DIY AI visibility audit: 12 checks — then sell your agency version with deeper query coverage, schema validation, and a prioritized sprint plan.

The proposal structure that survives legal and procurement #

1. Problem statement in the client's words (CTR, pipeline, competitor mentions)
2. Definition of AI visibility vs SEO in one short paragraph
3. Sample query bank results (anonymize competitor names if needed; keep screenshots)
4. Scope by phase with exit criteria for each phase
5. Pricing as ranges or fixed phase fees — never "unlimited GEO"
6. Measurement — citation rate on the bank, assisted conversions, branded search as a cross-check
7. Exclusions — no guaranteed placements; no buying junk links as "citation building"
8. Dependencies — access to Search Console, CMS, brand guidelines, subject-matter experts for E-E-A-T

Selling into an existing SEO retainer #

Upsell with a parallel workstream, not a rename of the same hours. Clients smell rebranded line items. I recommend a separate SOW or a clearly labeled addendum:

• Keep classic SEO KPIs on the original retainer
• Add AI visibility KPIs on the expansion
• Share a weekly ritual: 30 minutes on rankings/CTR, 30 minutes on citation scoreboard
• Cap net-new page volume until money URLs are extractable — restructuring often beats publishing for the first 60 days

Cold outreach vs warm expansion #

Channel	Offer	Proof asset
Warm SEO client	Audit + 30-day extractability sprint	Their own query bank sample
Warm but non-SEO client	Education webinar + paid audit	Anonymized vertical sample
Cold agency peer / white-label	White-label audit product	Fixed deliverable checklist + turnaround SLA
Cold end client	Diagnostic call + paid audit only	Public checklist + one vertical case pattern (no invented ROI)

My bias: warm expansion prints faster than cold "GEO agency" branding. Productize the audit first. Marketing polish second.

---

What is the best way to explain GEO to a client who only knows SEO? #

Explain GEO as SEO's cousin with a different scoreboard: instead of "rank and get the click," the goal is "get extracted and cited inside the answer." Same foundations — crawlability, trust, helpful content — different page shape and different proof. If you start with the acronym soup (GEO, AEO, AIO), you lose the room. Start with a side-by-side they already understand.

The dinner-table explanation #

Use this script almost verbatim:

"Google still has blue links. Your SEO program still fights for those. But a growing share of questions never produce a click — the answer appears in an AI Overview, in AI Mode, or inside ChatGPT or Perplexity. When that happens, the winner is not whoever ranked #1. The winner is whoever the answer names as the source. GEO / AEO work makes your pages easy for those systems to extract and trust. We are not throwing away SEO. We are adding a citation layer."

Then show three screenshots: a classic SERP, an AI Overview, and a ChatGPT or Perplexity answer. Point to citations. Ask: "Where is your brand?" Silence sells.

Translation table for SEO-native clients #

SEO concept they know	AI visibility equivalent	What changes in the work
Keyword target	Primary question + FAQ cluster	Briefs start with questions buyers ask AI, not only head terms
Meta description / title CTR	Lead-answer sentence under each H2	First 1–2 sentences must be extractable claims
Featured snippet chasing	Answer blocks, tables, FAQs	Structure for machines and skimmers
Backlink profile	Corroboration + entity consistency	Earn real mentions; stop selling junk links as GEO
Rank tracking	Citation rate on a query bank	Screenshots + logged mentions across surfaces
Content calendar by topic	Question-cluster publishing	Pillar/spoke mapped to question banks
Technical SEO	Technical SEO + schema/entity hygiene	FAQPage, Organization, Product/Service where relevant
E-E-A-T	E-E-A-T with attributable specifics	Named experts, dated sources, client-safe receipts

For the deeper content fork — what writers change day to day — send them GEO vs SEO: what actually changes in how you create content after the call. Do not make them read it before they understand the scoreboard.

GEO vs AEO vs AIO — only if they ask #

Keep this short. Clients care about outcomes, not taxonomy wars.

• GEO (Generative Engine Optimization): optimizing to be cited/recommended inside generative answers (ChatGPT, Perplexity, Gemini experiences, etc.).
• AEO (Answer Engine Optimization): optimizing for answer-shaped surfaces — featured answers, AI Overviews, voice-style Q&A.
• AIO (AI Overview / AI Optimization, depending on who is talking): in practice on this site, the work that makes pages ready for Google's AI features and answer extraction generally.

In proposals I often use AI visibility as the umbrella and mention GEO/AEO only in a footnote. Your client's CFO does not need the history of the terms.

Analogies that land (and one that fails) #

Works: "SEO is shelf placement in the store. AI visibility is getting named by the clerk when someone asks what to buy."

Works: "Rankings are your billboard on the highway. Citations are the recommendation your friend texts when someone asks for a vendor."

Fails: "SEO is dead, GEO is the future." You will trigger defensiveness from every SEO lead in the room — and you will be wrong. See the keep/drop/add argument in AI Visibility vs Traditional SEO.

A 10-minute live demo agenda #

1. Minute 0–2: Ask the client for five questions their buyers actually ask before purchasing.
2. Minute 2–6: Run those questions live across two AI surfaces + Google.
3. Minute 6–8: Show one competitor page that got cited — highlight answer-first structure, tables, FAQ, clear entity.
4. Minute 8–10: Show their closest money URL — highlight buried answers, thin FAQ, missing schema, vague claims.
5. Close: "The fix is not more blog posts first. It is extractability on these URLs, then a question bank we operate monthly."

Objections you will hear in this explanation #

Objection	Response
"Our rankings are fine."	"Rankings and citations are different KPIs. Let's check the query bank."
"AI just steals our content."	"Some clicks compress. Being the named source is how you stay in the shortlist when clicks shrink."
"Isn't this just featured snippets again?"	"Related skill, bigger surface area — ChatGPT, Perplexity, AI Mode, not only position zero."
"Can we wait a year?"	"You can. Competitors who become the default cited brand will be harder to displace later."
"We already use AI to write content."	"Generating drafts is not the same as designing pages for extraction and trust."

Teaching their content team without a mutiny #

Writers fear "write for robots." Reframe: write the answer a human would screenshot, then support it with proof. The AI visibility content strategy post is a good handoff document for in-house editors. Pair it with a one-page internal standard:

• H2 = a real question
• First sentence = direct answer in bold
• Include one structured element (table, list, steps)
• FAQ section with real questions
• Claims sourced or hedged
• Internal links only to live URLs

Models in the workflow (Claude Sonnet 5 / Claude Opus 4.8 for harder pages, GPT-5.5 / GPT-5.4 mini for outlines, Gemini 3.1 Pro / Gemini 3.5 Flash for research synthesis, Llama 4 for cost-sensitive loops) accelerate drafts. Humans own opinions, receipts, and client-safe specifics. Say that out loud so clients do not think you are selling unreviewed AI mush.

---

How do I package and price AI visibility services? #

Package AI visibility as three productized tiers — Audit, Foundation Sprint, and Citation Retainer — and price them as market-observed ranges tied to scope, not as a mysterious "GEO multiplier" on SEO hours. Buyers understand phases. They do not understand "strategic AI presence." Put deliverables, cadence, and exit criteria on one page.

Recommended offer architecture #

Tier A — AI Visibility Audit (productized, entry) #

Job: baseline + prioritized backlog your client can buy even if they do not retain you yet.

Element	Spec
Query bank	25–50 buyer questions, tagged by funnel stage
Multi-surface check	Google AI Overviews/AI Mode, ChatGPT, Perplexity (document date + model/UI notes)
On-site extractability review	Top 10–20 money URLs
Technical / schema / entity pass	Crawl issues, FAQPage/Organization/Product gaps, entity consistency
Competitive citation sample	3–5 competitors on the same bank
Deliverable	Scorecard + 30/60/90 backlog + optional SOW for Tier B
Timeline	1–3 weeks depending on site size

Pricing I have seen in mid-2026 (observations, not a price list you must copy): roughly $2.5k–$8k for SMB / single-location service businesses; $8k–$20k+ for multi-location, ecommerce, or complex B2B sites. Complexity drivers: number of URLs, locales, product catalog size, and how many AI surfaces you include. Hedge hard — your market, brand, and delivery costs differ.

Tier B — Foundation Sprint (30–60 days) #

Job: make money URLs extractable before you scale publishing.

Element	Spec
Page rewrites	Answer-first H2s, tables, FAQs on priority URLs
Schema implementation	Valid JSON-LD where it matches visible content
Entity cleanup	About/Organization consistency, author pages, sameAs hygiene
Internal linking	Pillar/spoke paths to commercial URLs
Measurement setup	Citation log template + reporting cadence
Exit criteria	Priority URL checklist complete; query bank re-tested

Pricing I have seen: often $6k–$25k as a fixed sprint, or a $4k–$12k/month intensive for 1–2 months. Ecommerce feed/schema-heavy work can run higher. Again: market observation, not a promise.

Tier C — Citation Retainer (ongoing) #

Job: operate the question bank, ship dual-purpose content, re-test citations, protect commercial URLs.

Monthly block	Example allocation
Query bank ops	Biweekly re-checks on priority questions
Content	2–6 question-cluster pages or major refreshes (scope-bound)
Technical hygiene	Schema QA, crawl fixes, template improvements
Reporting	Citation rate, notes on surface changes, next sprint picks
SEO coordination	Keep classic SEO work from colliding with AI visibility edits

Pricing I have seen: retainers commonly land around $3k–$10k/month for focused SMB programs; $10k–$25k+/month when you are running multi-market content ops plus engineering support. White-label delivery to other agencies often prices lower per end client but needs stricter SOPs. Treat these as ranges I have observed across conversations and proposals in the market — not guaranteed benchmarks and not named-client ROI.

Packaging rules that prevent scope chaos #

1. Cap questions in the bank. 25–50 active questions beats 500 ignored ones.
2. Cap URLs in the sprint. Top commercial + top informational money pages first.
3. Separate SEO hours from AI visibility hours in the invoice narrative.
4. Write exclusions. No guaranteed Overview placement. No link buying. No "we control ChatGPT."
5. Define revision rounds on rewrites.
6. Require SME access for E-E-A-T — or the content will be generic and fail citation tests.

Good / Better / Best menu (one-pager friendly) #

	Good	Better	Best
Entry	Audit only	Audit + Foundation Sprint	Audit + Sprint + 6-month retainer
Surfaces tracked	Google + one AI chat	Google + ChatGPT + Perplexity	Multi-surface + competitor panel
Content	Recommendations only	Rewrites on top URLs	Rewrites + monthly cluster publishing
Engineering	Issues list	Schema implemented	Schema + template-level extractability
Reporting	Scorecard	Scorecard + monthly	Biweekly ops + quarterly strategy

Clients like menus. Engineers like caps. Give both.

Pricing psychology for SEO agencies #

• Anchor on risk, not on hours. Hours invite comparison to cheap freelancers. Risk framing: "informational discovery is moving into answers; we measure and repair citation gaps."
• Phase fees beat blended mystery retainers for net-new buyers.
• Retainers beat phases once the scoreboard exists and the client sees movement — or at least a clear operating rhythm.
• Do not discount monitoring-only. Monitoring without fixes trains clients to hate the dashboard.
• White-label: sell deliverable clarity; protect your brand with anonymized methods if needed.

Sample pricing narrative (safe language) #

"For a single-location service business with a focused site, I've seen AI visibility audits quoted in the low-to-mid thousands, foundation sprints in the high four to low five figures, and retainers that sit beside SEO rather than replacing it. Multi-location and ecommerce programs run higher because entity and catalog complexity rises. We will quote fixed phase fees after we see your URL list and query bank size — not before."

That paragraph is honest. It does not invent a case study ROI. Use it.

Bundling with SEO without confusing the P&L #

Approach	Pros	Cons
Separate SKUs	Clean KPIs, easier upsell math	Two invoices / two conversations
Combined "Search + AI Visibility" retainer	Simple for client	Harder to defend scope; easy to under-deliver
Audit as paid diagnostic before retainer	Filters tire-kickers	Longer sales cycle
Performance bonus on citation rate	Aligns incentives	Attribution fights; avoid unless measurement is mature

My default recommendation for most agencies: paid audit → sprint → retainer, with SEO kept as a sibling SKU.

What to put in the price calculator (internal) #

• Number of priority URLs
• Number of locales / languages
• Catalog size (SKU count) if ecommerce
• Who implements schema (you vs client eng)
• Content production (you write vs you brief their writers)
• Meeting load (weekly vs biweekly)
• Legal/compliance review cycles (YMYL verticals)

If you ignore YMYL review time, you will lose money on healthcare and finance retainers.

---

Objection handling: the scripts that keep deals alive #

Most AI visibility deals die on four objections — guaranteed results, "SEO is fine," budget collision, and fear of AI writing slop — and each has a calm, specific answer. Memorize these. Do not argue philosophy.

"Can you guarantee we show up in ChatGPT / AI Overviews?" #

No. Decline guarantees. Say:

"No ethical operator can guarantee a specific AI surface will cite you on a timeline. Models, UI, and Overview triggers change. What we guarantee is the operating system: a query bank, extractable pages, valid structured data where appropriate, corroboration work, and a measurement cadence. That is how citation rate moves over time."

If a prospect demands guarantees, they are a bad-fit client. Let a competitor take the lawsuit risk.

"Our SEO is working — why add this?" #

"If commercial pages and brand search are healthy, great — keep funding that. AI visibility is about the questions that used to send informational clicks and now resolve inside answers. We will show you the query bank sample. If you are already cited everywhere that matters, you do not need us yet."

Then show the sample. Either they have a gap or they do not.

"We do not have budget beyond SEO." #

"Then start with a paid audit and a narrow foundation sprint on the top five money URLs. Do not buy a bloated retainer. Reallocate from low-yield informational volume if needed — fewer, better pages often beats more posts that never get cited."

Help them cut waste. Do not shame them for protecting SEO spend.

"Isn't this just AI content spam?" #

"If someone ships unreviewed model drafts with no receipts, yes — that fails. Our standard is answer-first structure, human expertise, sourced or hedged claims, and pages a subject-matter expert would defend on a sales call. Tools like Claude Opus 4.8 or Claude Sonnet 5 can accelerate drafts. They do not replace your experts."

"Our competitor hired an 'AI SEO' tool — should we?" #

"Buy software after the ritual exists. Dashboards that narrate your absence without a fix backlog are expensive anxiety. We install the query bank and extractability work first; tools second."

"Will this hurt our rankings?" #

"Done correctly, answer-first structure, clearer entities, and better FAQs usually help traditional SEO. It hurts when teams ship thin near-duplicates or break templates with sloppy schema. We coordinate with your SEO workstream and protect commercial URLs."

Point them to the keep/drop/add logic in AI Visibility vs Traditional SEO if they want depth.

---

The sales process: from first call to signed retainer #

Run a diagnostic-led process: educate → paid audit → sprint → retainer. Skip the giant strategy theater. Agencies that sell six-week "discovery" before any artifact create distrust. Ship a scorecard early.

Stage map #

Stage	Goal	Artifact	Exit
1. Discovery (30–45 min)	Confirm symptom + fit	Call notes + sample questions	Paid audit proposal
2. Paid audit	Baseline truth	Scorecard + backlog	Sprint SOW or polite no
3. Foundation sprint	Extractability on money URLs	Before/after pages + re-test	Retainer decision
4. Retainer	Ongoing citation ops	Monthly scoreboard	Renewal on operating value

Discovery call checklist #

• What does pipeline from organic look like vs 12 months ago?
• Which queries used to convert that feel softer now?
• Who is the buyer, and do they research with AI tools?
• What CMS and engineering access exist?
• Is there an in-house SEO lead who will feel threatened? (Plan the joint narrative.)
• Any YMYL constraints?
• Success definition in their words (demos, calls, stores visits, online revenue)

Red flags to walk away from #

• Demands guaranteed AI Overview placement
• Wants you to block all AI crawlers while also "winning GEO"
• Refuses any measurement beyond vanity screenshots
• Expects you to fabricate case-study ROI for their deck
• Will not give SME access for expert content
• Wants link schemes rebranded as citation building

Walking away is packaging. Bad clients destroy delivery margins and case-study quality.

---

What an agency-grade AI visibility audit should include #

Your audit is the productized wedge — make it boringly clear, visually scannable, and impossible to confuse with a generic SEO audit PDF. Clients should finish reading knowing exactly what is broken, what to fix first, and what it costs to continue with you.

Minimum viable audit sections #

1. Executive summary — three findings, three priorities, one recommended path
2. Query bank results — table with question, surface, cited brands, your status (cited / mentioned / absent)
3. URL extractability review — money pages scored on answer-first, structure, FAQ, schema, entity clarity
4. Technical & crawler notes — indexation, robots / AI crawler posture, performance on key templates
5. Entity & corroboration — knowledge consistency, about page, authors, third-party mentions
6. Competitive panel — why someone else gets named
7. 30/60/90 backlog — effort vs impact
8. Measurement plan — how you will prove progress without fake ROI
9. Commercial next step — sprint scope and price range

For a client-friendly checklist you can share pre-sale, align your scoring with the DIY AI visibility audit and deepen it with proprietary query coverage.

Scoring rubric (example) #

Dimension	0	1	2
Lead answer present	Buried or missing	Present but weak	Clear, bold, specific
Structured elements	None	Lists only	Tables/steps/comparisons
FAQ quality	None / fluff	Some real Qs	6–10 real Qs answered tightly
Schema match	None / invalid	Partial	Valid and matches visible content
Entity clarity	Confusing	Partial	Consistent across key URLs
Corroboration	Invisible off-site	Thin	Credible mentions exist

Sum scores per URL. Priority = low score × high commercial value.

How to present without overwhelming #

• Lead with the scorecard, not the crawl dump
• Show five example questions live if you can
• Put the full URL appendix behind the summary
• Always end on a decision: sprint yes/no

---

Deliverables your retainer must ship every month #

Retainers die when the only deliverable is a PDF of screenshots. Ship operating artifacts. Here is a monthly pack I recommend agencies standardize.

Monthly deliverable pack #

Artifact	Purpose
Updated citation log	Trend citation rate on the bank
Shipped pages / refreshes	Proof of extractability work
Schema / technical change log	Show engineering progress
Query candidates backlog	Next month's brief list
Risk notes	Surface volatility, crawler changes, competitor moves
Ask list for client SME	Keep E-E-A-T fed

Cadence that fits most SMB retainers #

• Week 1: query bank re-check + prioritization
• Week 2: content/schema production
• Week 3: QA, publish, internal links
• Week 4: report + next-month plan

Biweekly clients can compress weeks 1–2 and 3–4. Weekly meetings are usually overkill unless the account is enterprise.

Pair retainer work with a public execution checklist #

When clients ask "what are we doing this month," a concrete checklist beats vibes. Point account managers at the GEO checklist: 12 things to do this month as an education aid — then run your tighter, client-specific version inside the retainer.

---

Measurement and ROI conversations without inventing numbers #

Demonstrate progress with citation rate, assisted conversions, and branded search cross-checks — never with invented "300% AI ROI" slides. Procurement teams are getting smarter. Fake precision will burn your agency.

Primary metrics #

Metric	How to capture	Why it matters
Citation rate	% of query bank where brand is cited/mentioned	Core AI visibility KPI
Share of answer	How often you appear vs named competitors	Competitive framing
Extractability score	Rubric on priority URLs	Leading indicator
Assisted conversions	CRM notes / UTMs / "how did you hear"	Business bridge
Branded search	Search Console brand queries	Proxy for recommendation spillover
Classic SEO health	Rankings/CTR on commercial URLs	Prove you did not torch SEO

Secondary metrics (use carefully) #

• Referral traffic from AI products when identifiable
• Engagement on answer-first pages (scroll, CTA)
• Newsletter/demo conversions from citation-driven journeys (hard to attribute — say so)

ROI narrative that stays honest #

"We do not claim a universal ROI multiple for AI visibility work. What we can show is movement on citation rate for the questions that influence your buyers, improvements on extractability scores for money URLs, and any assisted pipeline your team can verify in CRM. Estimates vary by vertical and sales cycle. If someone promises a fixed return in thirty days, treat that as marketing, not measurement."

Reporting slide order that renews contracts #

1. Citation rate trend (sparkline or simple table)
2. Wins (newly cited questions) and losses (surfaces that changed)
3. Work shipped (URLs, schema, content)
4. Business notes (calls/demos influenced — only if real)
5. Next month priorities
6. Risks / dependencies

If you cannot fill slide 3, you are not operating — you are spectating.

---

Team roles: who does what inside the agency #

AI visibility delivery fails when it is "everyone's side hustle." Assign owners. Even a five-person agency can productize with clear RACI.

Role	Owns	Does not own
Strategist / account lead	Query bank, priorities, client narrative	Writing every page
Content lead	Answer-first briefs and QA	Guaranteeing citations
Technical SEO / eng	Schema, crawl, templates	Vague "AI strategy" decks
Researcher	Sources, competitor panels, claim hygiene	Publishing unreviewed drafts
Analyst	Citation log, reporting	Inventing ROI

Using models inside delivery (say it in the SOW) #

Be transparent with clients:

• Draft acceleration with Claude Sonnet 5 / Claude Opus 4.8
• Outline and FAQ expansion with GPT-5.5 / GPT-5.4 mini
• Research synthesis with Gemini 3.1 Pro / Gemini 3.5 Flash
• Optional cost-sensitive loops with Llama 4
• Human review mandatory before publish
• No fabricated testimonials or stats

Clients who forbid AI tooling entirely may still buy audits; they are often poor fits for content retainers at agency margins.

---

Positioning your agency in a crowded "AI SEO" market #

Win on specificity of method and honesty of measurement, not on who shouts "GEO" loudest. The market is noisy. Your differentiator is a productized audit, a clear sprint, and a scoreboard — plus the judgment to keep SEO foundations intact.

Positioning angles that work #

1. SEO shop adding a second optimization layer — trustworthy for existing clients
2. AI visibility studio partnering with SEO agencies (white-label) — good if you lack sales distribution
3. Vertical specialist (legal, home services, B2B SaaS) — higher close rates, clearer proof patterns
4. Technical + content hybrid — schema/entity strength plus editorial standards

Positioning angles that usually fail #

• "SEO is dead, hire us"
• Guaranteed placement claims
• Pure tool reseller with no delivery team
• Rebranded link building
• Vague "share of model" language with no query bank

Should you rename the agency? #

Usually no — not first. Add a service line and proof. Rename only if your brand is trapped in a dead category and sales already stall on first impression. Most SEO agencies close faster by saying "Search + AI Visibility" than by throwing away brand equity for a trendy GEO name.

Website and sales collateral updates (practical) #

• Service page: AI visibility / GEO / AEO with packages
• Case patterns: before/after extractability and citation logs (with permission)
• FAQ on your service page matching buyer questions
• Sales one-pager with the Good/Better/Best menu
• Internal wiki with banned claims list (no guarantees, no fake ROI)

---

White-label and partner motions #

If you lack distribution, sell delivery to agencies that already own SEO relationships. If you own distribution but lack technical depth, buy white-label fulfillment — with SOPs so quality does not collapse.

White-label product specs #

Item	Recommendation
SKU	Audit + Sprint + Retainer modules
Turnaround	Audit in 10 business days for standard scope
Branding	Partner-facing templates; your methods anonymized if needed
Communication	Partner owns client; you own delivery Slack with partner
QA bar	Same rubric as direct clients
Pricing	Partner margin built in; do not race to the bottom

Partner enablement kit #

• 10-minute GEO explainer script
• Objection card
• Sample anonymized scorecard
• Pricing bands (ranges, hedged)
• Scope caps checklist

Train partners not to promise magic. Your brand still takes the blast radius when their AE oversells.

---

30/60/90 day plan for agencies launching the offer #

Do not wait for a perfect rebrand. Ship the audit product in thirty days.

Days 1–30 — Productize #

• Lock Tier A/B/C scopes
• Build scorecard template
• Pick one vertical for a sample query bank
• Run 2–3 free or discounted audits for friendly clients in exchange for testimonials (real ones)
• Write the service page draft with answer-first structure (practice what you sell)
• Train AEs on the dinner-table GEO script

Days 31–60 — Sell and deliver #

• Launch paid audits to existing SEO book
• Deliver 2–4 audits
• Convert at least one sprint
• Formalize monthly reporting pack
• Kill any marketing copy that implies guarantees

Days 61–90 — Systemize #

• Create SOP library (briefs, schema QA, citation log)
• Decide white-label yes/no
• Build a small claim/source checklist for writers
• Review margins — raise prices if delivery quality is high and win rate is strong
• Add case patterns (process + metrics), still without inventing ROI

---

Content and page standards you should sell as non-negotiables #

Your delivery quality bar is part of packaging. If clients insist on thin posts for volume vanity, price it as a separate cheap SKU or decline. Citation work needs standards.

Non-negotiables #

• Question-shaped H2s with bold lead answers
• One structured element per major section
• Real FAQ section (not fluff)
• Claims sourced or explicitly hedged
• Schema only when it matches visible content
• Internal links only to live pages
• Expert review on YMYL

Hand clients the AI visibility content strategy as the shared standard. Use the GEO checklist for monthly hygiene. Use GEO vs SEO content changes when training writers who still brief like it is 2019.

---

Competitive teardown: how to talk about "AI SEO" vendors #

Acknowledge tools; sell judgment and execution. Prospects will compare you to platforms. Be ready.

Prospect belief	Your response
"The tool will fix GEO."	"Tools monitor and sometimes suggest. They do not rewrite your money pages or fix entity contradictions."
"They guarantee mentions."	"Ask what happens in the contract if Google changes Overview triggers next month."
"They are cheaper."	"Compare deliverables: query bank size, rewrite ownership, schema implementation, SME process."
"We can DIY with ChatGPT."	"You can DIY an audit checklist. Operating it monthly across surfaces while shipping extractable pages is the hard part."

Never trash a specific vendor with unverified claims. Compete on clarity.

---

Proposal appendix: sample scope language you can adapt #

Steal this structure. Rewrite in your voice. Keep the exclusions.

Sample scope — Foundation Sprint #

In scope:

• Prioritize up to 15 URLs from the audit backlog
• Rewrite or restructure pages for answer-first extraction
• Implement agreed schema types matching visible content
• Build or refine a 40-question citation bank
• Re-test the bank at sprint end across agreed surfaces
• Provide a next-phase retainer recommendation

Out of scope:

• Guaranteed citations or Overview placements
• Purchased links or fake reviews
• Full site redesign
• Unlimited page production
• Legal advice; client counsel reviews YMYL claims

Client dependencies:

• CMS access, Search Console access, analytics access
• SME interviews within five business days of request
• Brand guidelines and banned-claim list

Sample success criteria #

• Extractability rubric improved on ≥80% of sprint URLs
• Citation log established with baseline + end-of-sprint measurements
• Schema validation errors resolved on targeted templates
• Joint decision meeting on retainer scope

Notice what is missing: a fabricated revenue multiple.

---

Common packaging mistakes (and fixes) #

Mistake	Why it hurts	Fix
Renaming SEO hours as GEO	Client feels bait-and-switch	Separate SKUs and KPIs
Unlimited content promises	Margins die; quality dies	Cap pages; prioritize money URLs
Monitoring-only retainers	No progress story	Bundle fixes + measurement
No exclusions	Scope creep + legal risk	Write them in every SOW
One global price	Wrong-fit clients and underpricing	Calculator by URLs/locales/catalog
Ignoring SEO team politics	Internal blockers	Joint kickoff with SEO lead
Overusing acronyms	Confusion, stalled deals	Lead with outcomes, footnote terms
Fake case ROI	Trust destruction	Use citation and process proof

---

Vertical notes: how packaging shifts #

Same tiers, different weights. Do not force a SaaS playbook onto a plumber.

Local services #

• Heavier on Google surfaces, reviews, location entity clarity
• Smaller query banks, faster sprints
• Retainer content often = service page FAQs + city/service hybrids done carefully (avoid doorway spam)

B2B SaaS #

• Heavier on comparison questions and integration entities
• Longer sales cycles → assisted conversion tracking matters more
• Docs and changelog pages can be citation assets if public and clear

Ecommerce #

• Product schema, feed hygiene, category extractability
• Query banks include "best for / vs / alternative" patterns
• Engineering time rises — price it

YMYL (health, finance, legal) #

• SME and compliance gates in every phase
• Slower publishing, higher prices
• Stronger author/entity requirements

---

Discovery workshop: run this with every serious prospect #

A ninety-minute workshop beats three vague strategy calls. Use it after the paid audit is scoped — or as a premium discovery for enterprise accounts. The goal is shared language and a ranked question list, not theater.

Workshop agenda #

Block	Minutes	Output
Business model and buyers	15	Who asks what before they buy
Money URL inventory	15	Top commercial + informational pages
Live AI query sampling	25	Raw citation absences
Competitor panel	15	Why others get named
KPI negotiation	10	Citation + business metrics you will report
Package recommendation	10	Tier A/B/C decision

Facilitation tips #

• Bring a prepared starter bank of 20 questions so the room is not blank
• Force specificity: "best CRM for dentists" beats "CRM software"
• Capture exact phrasing from sales reps — they hear real buyer language
• Park legal debates in a parking lot; do not burn the workshop on compliance philosophy
• End with a written decision: audit only, sprint, or not a fit

Workshop artifacts to leave behind #

1. Ranked question list (draft bank)
2. Money URL list with owners
3. Screenshot pack from the live session
4. Recommended package one-pager
5. Open questions / dependencies

Prospects who refuse a workshop and demand a giant retainer on day one are often buying a logo, not an operating system. Price risk accordingly — or decline.

---

Hiring and training: building an AI visibility pod #

You do not need a twenty-person "GEO department." You need a pod that can audit, rewrite, implement schema, and report. Start small. Hire for judgment and editorial standards before you hire for slide design.

Pod shapes by agency size #

Agency size	Pod shape	Notes
2–5 people	Strategist + writer + fractional technical SEO	Founder often sells
6–20 people	Dedicated AI visibility lead + content + tech support from SEO team	Shared analytics
20+	Pod per vertical or a central center-of-excellence	Watch for process bloat

Skills to screen for #

• Can write an answer-first section without fluff
• Understands schema matching visible content (not spam markup)
• Comfortable saying "I do not know" when a model output invents a stat
• Can run a query bank without turning it into a vanity science project
• Diplomacy with SEO leads who feel threatened

Thirty-day internal training plan #

Week 1: Read the keep/drop/add pillar and practice the dinner-table GEO script on each other.

Week 2: Each person runs a 15-question bank on a friendly client site and scores five URLs with the rubric.

Week 3: Pair write one answer-first page; peer review against the non-negotiables list.

Week 4: Dry-run an audit readout; record it; fix pacing and jargon.

Internal enablement compounds. A single trained AE who can run the live demo will outsell a beautiful deck that nobody can present.

---

Legal and compliance language for SOWs #

Put humility in the contract. AI surfaces change. Your lawyers will thank you. Your future self will thank you when a client forwards a ChatGPT screenshot from a Tuesday UI that disappeared on Wednesday.

Clauses worth including (plain-English intent) #

• No guaranteed placements in any AI product or Google AI feature
• Best-effort measurement across surfaces that may change without notice
• Client represents that SME-provided facts are accurate
• YMYL review remains client counsel's responsibility
• Model-assisted drafting disclosed; human review required before publish
• Termination for convenience with payment for work completed
• Portfolio rights optional — ask permission before publishing case patterns

I am not your attorney. Have counsel adapt this. The point is cultural: your sales team should not verbally promise what the SOW refuses.

Claims hygiene inside delivery #

Claim type	Rule
Statistics	Need dated source or hedge as estimate
Model capabilities	Use current names only; verify before publish
Competitor weaknesses	Stick to observable page/citation facts
Client results	Only with permission; no invented ROI
"Everyone knows" statements	Delete or replace with evidence

This is also how you avoid banned fluff in client deliverables — the same standards you use on williamspurlock.com.

---

Pricing scenarios (worked examples) #

These scenarios are illustrative packaging math from patterns I have seen — not quotes, not promises, and not named-client results. Use them to train AEs on scope thinking.

Scenario 1 — Single-location home services company #

• Site: ~40 URLs, WordPress, thin service pages
• Pain: "best near me" style questions cite national brands
• Package: Audit ($3k–$5k range I have seen) → Sprint on 8 service pages → $3.5k–$6k/month retainer for FAQ expansion + citation log
• Risk: doorway-page temptation — keep quality high, avoid spammy city clones

Scenario 2 — B2B SaaS with strong SEO blog #

• Site: large blog, solid rankings, soft demo assist from informational content
• Pain: comparison questions cite competitors inside Perplexity/ChatGPT
• Package: Audit on 40-question bank → Sprint on pricing, compare, and integration URLs → higher retainer for cluster publishing
• Pricing pattern I have seen: audit in the high four/low five figures; retainer often $8k–$15k/month depending on volume and eng support

Scenario 3 — Multi-location healthcare (YMYL) #

• Constraints: medical review on every page
• Package: slower sprint, higher price, explicit compliance gates
• Do not sell speed. Sell safety + extractability on service lines that already convert
• Expect fewer pages per month; price the review overhead

Scenario 4 — Ecommerce mid-catalog #

• Pain: category pages rank but AI answers recommend competitors' buying guides
• Package: technical/schema-heavy sprint + selective editorial guides
• Engineering share of budget rises; content-only pricing will underfund you

How to present scenarios without over-anchoring #

"These are packaging patterns, not your quote. Your fee depends on URL count, review cycles, and who implements schema. We price after the audit defines the backlog."

---

Onboarding checklist after the contract is signed #

Fast onboarding protects margins. Slow access kills sprints.

Day 0–3 #

• Collect CMS, Search Console, analytics, brand guidelines
• Confirm SME calendar holds
• Finalize query bank v1
• Freeze scope caps in writing

Day 4–10 #

• Complete baseline citation log
• Score money URLs
• Deliver kickoff readout (even if audit already done — align the working team)
• Lock sprint URL list

Day 11–30 #

• Ship first rewrites
• Implement schema on agreed templates
• Re-test a subset of the bank
• Show early operating rhythm in a short report

If access stalls past two weeks, escalate commercially. Free waiting is not a deliverable.

---

Renewal and expansion plays #

Renew on operating value: the bank is alive, pages shipped, citation rate discussed like adults. Expansion comes from new locales, new product lines, or white-label siblings — not from fear screenshots alone.

Renewal conversation script #

"Here is citation rate on the agreed bank at baseline vs now. Here is what we shipped. Here is what changed in AI surfaces that we adapted to. Here is the next quarter backlog. If the scoreboard and the shipping cadence still matter to pipeline, we continue. If your priority shifted entirely to paid media, we should resize or pause rather than pretend."

Honesty renews more often than panic.

Expansion triggers #

• New market / language
• New product family
• Merger/acquisition entity cleanup
• Ecommerce catalog growth
• Partner/reseller content needs
• Internal team wants enablement training

When to downsell #

If the client cannot supply SMEs, or leadership only wants vanity PDFs, downsell to a quarterly audit rather than a hollow monthly retainer. Protect your case-study quality.

---

Competitive intelligence without creepy theater #

Track competitors on the same query bank. Do not claim you know their "AI strategy" from vibes. Observable facts only.

What to log monthly #

• Which competitors are cited on priority questions
• Which of their URLs appear as sources
• Structural patterns on those URLs (FAQ, tables, authors)
• New corroboration (press, partner pages, docs)
• Changes in their schema footprints when visible

What not to do #

• Invent narratives about their internal roadmap
• Copy their content verbatim
• Obsess over one lucky ChatGPT mention
• Promise you will "dethrone" them in thirty days

Use competitor panels to prioritize your backlog. If they win on comparison pages with clean tables and dated sources, that is your sprint brief — not a reason to buy junk links.

---

Tooling stack: buy late, standardize early #

Standardize templates before you standardize SaaS. Your stack can be boring.

Need	Lightweight start	Later upgrade
Query bank	Sheet / Airtable	Dedicated monitoring tools
Citation log	Sheet with date/surface/result	Automated alerts
Content briefs	Doc template	Project management workflow
Schema QA	Rich results testing + validators	CI checks in client engineering
Reporting	Slide/doc pack	Dashboard + narrative

Model use inside the agency (recap for operators) #

• Claude Opus 4.8 / Claude Sonnet 5: harder rewrites, audit synthesis, strict editorial passes
• GPT-5.5 / GPT-5.4 mini: outlines, FAQ expansion, bulk restructuring suggestions
• Gemini 3.1 Pro / Gemini 3.5 Flash: research synthesis, multimodal checks on page screenshots
• Llama 4: cost-sensitive internal loops where policy allows

Whatever you use, the publish gate stays human. Clients are buying your standards.

---

Pitch deck outline (12 slides) #

If you need a deck, keep it short. The live query demo does the heavy lifting.

1. The symptom: rankings ≠ recommendations
2. Side-by-side: blue links vs AI answers
3. Definition: AI visibility in one slide
4. Translation table (SEO → citation work)
5. Sample query bank results (prospect-specific if possible)
6. Offer tiers: Audit / Sprint / Retainer
7. Deliverables calendar
8. Measurement (and what you refuse to fake)
9. Team + workflow (including model-assisted drafting disclosure)
10. Timeline 30/60/90
11. Pricing approach (ranges + next step quote)
12. CTA: paid audit

Remove any slide that exists only to sound impressive. Add screenshots until it hurts.

---

Internal FAQ for account executives #

Train AEs on these so they stop improvising dangerous promises.

What if the client asks for a free audit? #

Offer a free 10-question sample, not a full audit. Full audits are the product. Free full audits train the market to undervalue diagnosis.

What if they want us to block AI crawlers? #

Discuss tradeoffs explicitly. Blocking can limit training/use of content in some systems while also reducing the chance those systems cite you. Align on goals before touching robots rules. Do not make unilateral changes.

What if their SEO vendor feels threatened? #

Invite them into the kickoff. Position as a complementary layer. Share the keep/drop/add framing. If the other vendor badmouths citation work as fake, stick to the query bank evidence.

What if leadership only cares about sessions? #

Reframe to pipeline and shortlist inclusion. Offer a dual dashboard: classic SEO metrics + citation metrics. If they refuse a second KPI after education, they may not be ready.

What if a junior marketer wants 50 AI posts per month? #

Cap volume. Sell extractability on money URLs first. Volume without structure creates sludge and can dilute entity clarity.

---

Quality control checklist before any client publish #

Run this checklist on every page your AI visibility retainer ships.

• Primary question is obvious in title/H1/H2
• First answer appears in the first 1–2 sentences of the section
• At least one table, list, or step sequence exists
• FAQ questions are real buyer questions
• Every hard number has a source or a hedge
• No guaranteed-ranking or guaranteed-citation language
• Schema matches visible content
• Internal links resolve to live URLs
• Author/SME reviewed when required
• Page is crawlable and not trapped behind pointless gates
• Claims would survive a skeptical sales engineer reading them aloud

If a page fails three or more boxes, it is not ready — even if a model drafted it beautifully.

---

Closing the loop with SEO services you already sell #

Your SEO menu still matters: technical health, commercial landing pages, digital PR that earns real mentions, local presence where relevant. AI visibility packaging should reference those SKUs as siblings.

Existing SEO SKU	How AI visibility attaches
Technical SEO	Schema/entity and crawler posture expansions
Content retainer	Question-cluster briefs + answer-first standards
Digital PR	Corroboration for citations, not fake link blasts
Local SEO	Entity consistency + review/location clarity
Analytics retainer	Citation log alongside Search Console

Cross-train teams. The agencies that win will not be "SEO or GEO." They will be operators who can sell both without lying about either. For the strategic backbone your delivery team should share, keep AI Visibility vs Traditional SEO and GEO vs SEO content changes in the internal wiki.

---

Frequently Asked Questions #

What does an AI visibility audit look like for a client deliverable? #

An agency-grade AI visibility audit is a scorecard plus backlog: a 25–50 question citation bank across Google AI surfaces, ChatGPT, and Perplexity; extractability scores on money URLs; schema/entity findings; a competitive panel; and a 30/60/90 plan. It should read in fifteen minutes at the executive layer, with appendices for specialists. If the PDF is only a crawl export with "GEO" on the cover, it is not an AI visibility audit.

How do I demonstrate ROI of AI visibility work to clients? #

Demonstrate ROI with citation-rate movement on agreed buyer questions, extractability improvements on commercial URLs, and assisted pipeline your client can verify — not with invented percentage returns. Tie reporting to their CRM reality. Estimates vary by vertical and cycle length; as of mid-2026, treat branded search and sales anecdotes as cross-checks, not as sole proof. Refuse clients who demand a guaranteed revenue multiple up front.

Should I rebrand my SEO agency to include GEO and AEO services? #

Usually add a service line first; only rebrand if your current name actively blocks sales. Most firms close faster with "SEO + AI Visibility" positioning, proof from audits, and clear packages than with a full rename. Rebrand when category perception is stuck and you already have delivery proof — not as a substitute for a productized offer.

What is a fair price to charge for AI visibility optimization in 2026? #

From market conversations and proposals I have seen in mid-2026, paid audits often land roughly in the low thousands to low tens of thousands, foundation sprints in the high four to mid five figures, and monthly retainers commonly somewhere in the low-to-mid thousands up through higher five figures for complex programs — with wide variance by scope. These are observations, not universal rates. Price from URL count, locales, catalog complexity, engineering ownership, and YMYL review load. Hedge every public number; quote fixed fees after discovery.

How is selling GEO different from selling traditional SEO retainers? #

You sell a second scoreboard — citations inside answers — and a different content shape, while still protecting classic rankings. The discovery asset is a query bank sample, not only a rank report. Packages skew toward audit → extractability sprint → citation ops, because publishing volume alone underperforms when pages are not extractable. Measurement must include citation logs alongside Search Console.

What should I include in an AI visibility monthly retainer? #

Include query-bank operations, a capped number of answer-first pages or refreshes, schema/technical hygiene, citation reporting, and coordination with SEO so commercial URLs stay protected. Exclude guarantees and unlimited production. If the monthly pack is only screenshots, you are selling anxiety. Ship work and a scoreboard.

How do I handle clients who want guaranteed ChatGPT mentions? #

Say no, explain volatility, and reframe guarantees around operating system delivery: bank, extractability, schema hygiene, cadence, and measurement. If they insist, walk away. Guaranteed mention clauses create legal and reputational risk you cannot control when model UIs and Overview triggers change.

Can I white-label AI visibility services for other agencies? #

Yes — productize audits and sprints with SLAs, partner-facing templates, and strict rules against overpromising. White-label works when the partner owns the relationship and you own delivery quality. Train their AEs on the GEO explainer and objection scripts so they do not sell miracles your team cannot ship.

---

Get help packaging AI visibility for your book of business #

If you run an SEO practice and your clients are asking why informational CTR looks thinner while competitors show up inside Google AI Overviews, ChatGPT, and Perplexity, you do not need another acronym. You need a productized audit, a foundation sprint, and a retainer scoreboard your account team can sell without flinching.

I help agencies and consultants position AI visibility offers, build AI-visibility-ready site patterns, and pressure-test packaging so you are not inventing ROI slides to close. Bring your current SEO SKUs and two sample clients.

Book an AI visibility strategy call and we will map audit → sprint → retainer around the questions your buyers already ask — with honest measurement and no placement fairy tales.