CRO Audit of Our Own Services Page

A full conversion rate optimization audit of this site’s own AI SEO services page, run on September 13, 2026 with the same eleven-phase method AI Search Engine Optimization applies to client pages, and published in full: every element captured verbatim, every score with its reason, and the same ten dimensions scored on the largest competitor ranking for the term.

Site: aisearchengineoptimization.com Page: https://aisearchengineoptimization.com/ai-seo/ Audit date: September 13, 2026 Framework: page-cro v2.5 (11 phases, 12 dimensions, composite out of 120) Auditor: AI Search Engine Optimization, Rank Intelligence Engine Comparison page: NP Digital, “AEO / GEO Agency Services” (https://npdigital.com/solutions/earned-media/ai-search-engine-optimization/), scored on the ten page dimensions only

---

Phase 0: Input Confirmation

Phase 0 Completion Gate: all three items YES. The URL is the page itself, not the parent site.

---

Phase 1: Page Capture and Evidence Bank

1.1 Hero block (verbatim)

• Eyebrow: “AI SEARCH VISIBILITY, WITH A BUSINESS PURPOSE”
• H1: “AI SEO Services”
• Subheadline (H2): “Be part of the decision”
• Supporting copy: “Help prospective customers discover and consider your business when they ask Google, ChatGPT, Perplexity, Gemini or Google’s AI Overviews for a service recommendation. Our AI SEO services cover the AI generated answers that now sit above the classic list of links, as well as the links themselves. It is AI search engine optimization and traditional SEO delivered as one service.”
• Secondary paragraph: “Our proprietary Rank Intelligence Engine connects Phil Belleville’s SEO experience dating to 1995 with 25 specialized analysis and implementation capabilities, persuasive content, and WordPress implementation.”
• Hero CTA button: “Book a free strategy call” (to /book-a-call/)
• CTA microcopy: “30 minutes with Phil. Clear next steps. No obligation.”
• Secondary hero CTA: none in the hero body; header carries “Book a free call” (to /book-a-call/)
• Hero image: none. The right column is a dark card headed “RANK INTELLIGENCE ENGINE / From discovery to consideration.” with three numbered steps: “01 Be understood: Present a clear picture of your business and services.” “02 Build confidence: Give prospective buyers credible reasons to consider you.” “03 Invite action: Make the next step easy when someone reaches your site.”

1.2 Heading inventory (verbatim, in order)

• H2: Be understood
• H2: Build confidence
• H2: Invite action
• H2: Make your expertise easier to discover
  • H3: Reach the right questions
  • H3: Make your value clear
  • H3: Connect interest to inquiries
• H2: One system A complete view
  • H3: A stronger business identity
  • H3: A clearer competitive position
  • H3: A website ready for the visit
• H2: What an AI answer needs before it names your business
  • H3: Entity clarity: can the engines tell who you are?
  • H3: Google’s own record: the Knowledge Graph
  • H3: Expertise and trust the engines can verify
  • H3: Topical depth: does the site know its subject?
  • H3: The block an AI will quote
  • H3: Where the engines confirm you: the citation footprint
• H2: What each AI platform reads, and why one page cannot serve them all the same way
  • H3: Google AI Overviews and AI Mode
  • H3: ChatGPT and Microsoft Copilot
  • H3: Perplexity and Gemini
• H2: The work ends on your WordPress website, not in a document
  • H3: Pages built to the bar
  • H3: Structured data as a mirror of the page
  • H3: Measured by platform, question and date
• H2: For businesses ready to be considered
• H2: Visibility is the start Business value is the goal
• H2: Your questions Clear answers
• H2: Business perspectives on AI SEO Services (ten linked H3 post titles beneath it)
• H2: Give the right buyers more reasons to consider you

1.3 CTA inventory

Phone number: ABSENT on this page. No tel: link anywhere in the page HTML.

1.4 Trust signal inventory

1.5 Social proof quotes (verbatim) ABSENT on this page. No testimonial or review quote appears anywhere on /ai-seo/.

1.6 Form inventory ABSENT on this page. Every CTA links to /book-a-call/; no form is embedded on the audited page.

1.7 Friction elements visible in copy

• Pop-ups: none. Rendered at 1366 by 768 and 390 by 844 on September 13, 2026: zero modals, zero interstitials; the only fixed element is a 41 by 41 pixel scroll-to-top control at the bottom right.
• Required form fields: none on this page.
• Length to first CTA: within the first screen at both widths (hero CTA top at 685 pixels on desktop; “Book a free strategy call” visible above the fold on phone).
• Navigation distractions: 50 navigation links in the header, including two dropdown menus (Services, About, The AISEO Process).
• Explicit confirmation copy: “30 minutes with Phil. Clear next steps. No obligation.” beneath the hero CTA; “No obligation. A useful next step.” beneath the closing CTA.
• Page load: 1.0 second to load event on desktop, 0.5 seconds on phone (measured September 13, 2026).

1.8 SEO elements

• Title tag: “AI SEO Services for WordPress Websites | Rank Intelligence Engine”
• Meta description: “AI SEO services that get a WordPress business found in Google and named in ChatGPT, Perplexity, Gemini and AI Overviews, with every change verified live.”
• Target keyword inferred: “AI SEO services” (H1, URL /ai-seo/, first 100 words), with “AI SEO services for WordPress” as the registry Aux.
• Word count: 3,329 words of visible body text.

Phase 1 Completion Gate: all three items YES. Every ABSENT note is an absence found on the page, not an unchecked item.

---

Phase 2: Attention Competition and 5-Second Clarity Verdicts

2.1 Attention competition test

1. Does the H1 “AI SEO Services” stop the scroll on its own? NO. It is a category label. It names the service the visitor searched for, which confirms the page is relevant, but it carries no outcome, no number and no recognition trigger.
2. Is the first section more interesting to read than to skip? MARGINAL. The supporting copy names the platforms (“ask Google, ChatGPT, Perplexity, Gemini or Google’s AI Overviews for a service recommendation”) which mirrors the prospect’s situation, but the sentence that follows (“It is AI search engine optimization and traditional SEO delivered as one service”) is a definition, not a reason to keep reading.
3. Dropped onto a busy feed, would the right prospect pause on the H1? NO. “AI SEO Services” would read as an ad category.

Attention Verdict: MARGINAL. The hero copy beneath the H1 does the work the H1 should do; the H1 itself is inert. Logged as a Critical Fix candidate.

2.2 Five-second clarity test

1. What is this? YES. AI SEO services for a WordPress business, stated in the H1, title and first sentence.
2. Who is it for? PARTIAL. “Your business” and “WordPress implementation” appear in the hero; the explicit fit statement (“businesses with a real offer, relevant experience, and a WordPress website”) sits far down the page in “For businesses ready to be considered”.
3. Why should I care? YES. “discover and consider your business when they ask Google, ChatGPT, Perplexity, Gemini or Google’s AI Overviews for a service recommendation” states the stake in the second sentence.
4. Is the most important message above the fold? YES on desktop and phone: H1, subhead, supporting copy, CTA and microcopy all render in the first screen.
5. One clear next step? YES. Every CTA on the page goes to /book-a-call/; there is no competing form, phone or chat.
6. Local and specific, or generic? SPECIFIC to this business: “Rank Intelligence Engine”, “Phil Belleville”, “since 1995”, “25 specialized analysis and implementation capabilities” are all named in the hero.

Clarity Verdict: PASS. A visitor knows what the page is, why it matters and what to do within five seconds; the who-it-is-for statement arrives late.

Last-Click Checkpoint: ENDS-THE-SEARCH. The above-fold content answers the mapped query (what AI SEO services are and what they do) rather than withholding, and every logical next step has an on-page path: price and scope (FAQ “written scope and price”, “month to month”), proof (case studies link), related questions (eight FAQs and ten linked perspectives), booking (three CTAs). Nothing forces a return to the results.

Phase 2 Completion Gate: all three items YES. The MARGINAL attention verdict is logged as a Critical Fix candidate for Phase 7.

---

Phase 3: VOC Research and Phrase Bank

Step A, Review Corpus detection: no review_corpus file in /mnt/user-data/uploads/ or this session. CORPUS_PRESENT = NO. Step A2, Prospect-Psychology Brief detection: no Gap Brief or Epiphany Intelligence Brief for this business in uploads or this session. BRIEF_PRESENT = NO. Fall-through: four web searches executed on September 13, 2026.

VOC Phrase Bank (15 rows, all verbatim)

Phase 3 Completion Gate: all five items YES. Four searches executed, 15 verbatim phrases with source URLs, bank delivered before any rewrite.

---

Phase 4: 12-Dimension Page Analysis

Dimension 1: Attention and Hook Quality

• Finding: The H1 “AI SEO Services” (1.1) is a category label. It requires EVALUATE, not FIRE: the prospect has to decide whether “AI SEO services” describes what they need. No question-form headline, which is correct. The recognition work is done one line lower by “when they ask Google, ChatGPT, Perplexity, Gemini or Google’s AI Overviews for a service recommendation”, which mirrors VOC phrase 11 (“invisible when someone asks ChatGPT, Gemini, or Perplexity for recommendations”).
• Severity: HIGH.
• Fix direction: put the outcome the prospect already wants into the H1 while keeping the Main “AI SEO Services” intact for containment.

Dimension 2: Value Proposition Clarity

• Finding: The first 100 words name the outcome (“discover and consider your business”), the situation (the five platforms) and the difference (“AI search engine optimization and traditional SEO delivered as one service”). It is outcome-led and in the buyer’s frame. It weakens at “It is AI search engine optimization and traditional SEO delivered as one service”, a definition sentence, and the second paragraph (“25 specialized analysis and implementation capabilities”) is about the vendor rather than the buyer.
• Severity: MEDIUM.
• Fix direction: keep the first two sentences, move the capabilities sentence below the CTA.

Dimension 3: Headline and Hero Section

• Finding: The H1 carries no specificity (no timeframe, number or outcome). The subhead “Be part of the decision” (1.1) supports the concept but is abstract. There is no hero image; the three-step card (“Be understood / Build confidence / Invite action”) is clean and does not distract, but it repeats the abstract register of the subhead. The trust strip beneath it (“Since 1995”, “WordPress”, “Founder led”) is the most specific element in the first screen and sits below the CTA.
• Severity: HIGH.
• Fix direction: specificity into the H1 and subhead; lift the trust strip to sit beside or directly under the CTA.

Dimension 4: CTA Placement, Copy and Hierarchy

• Finding: Primary CTA “Book a free strategy call” is above the fold at both widths, communicates value (“free”, “strategy”) and carries the strongest microcopy on the page: “30 minutes with Phil. Clear next steps. No obligation.” The header CTA “Book a free call” is persistent. The closing CTA “Book your free strategy call” repeats it. The gap: across 3,329 words there is no CTA between the hero (first screen) and the closing section; the proof section (“Visibility is the start”) and the FAQ have no CTA adjacent to them. No phone number and no click-to-call, which is defensible for a national B2B booking flow but removes the alternative path a hesitant visitor takes.
• Severity: MEDIUM.
• Fix direction: repeat the booking CTA after the accountability section and after the FAQ; keep booking as the single path.

Dimension 5: Trust Signals and Social Proof

• Finding: From the 1.4 inventory: no star rating, no named testimonial, no client logos, no association badges, no press, zero images. The page’s proof is three things: “Since 1995”, “Founder led”, and one paragraph pointing to eight case studies that the page itself says “do not establish AI citation or revenue gains”. That honesty is a strength for AI citation and a weakness for a nervous buyer whose language is VOC 5 (“absolutely zero positive results”) and VOC 8 (“zero results, zero effort and zero accountability”). No proof sits before or after any CTA.
• Severity: CRITICAL.
• Fix direction: put one attributed, specific proof element beside each CTA; surface at least one countable result from the eight case studies on this page.

Dimension 6: Objection Handling and FAQ Quality

• Finding: Eight FAQs, each 3 to 5 sentences, entity-rich, answering real objections: what the service is, how it differs, whether AEO and GEO are included, which platforms are measured, implementation, the guarantee question (“No. Each platform controls its own answers”), and how to begin. Pricing is handled by process (“written scope and price”, “month to month”) with no range, which leaves VOC 9 (“they’re expensive”) unanswered. The risk objection (VOC 10, “more harm than good”) is answered implicitly by “verified live” but not named. No FAQ ends with a soft CTA. Eight is at the floor of the 8-to-20 standard.
• Severity: MEDIUM.
• Fix direction: add three FAQs (price range or how price is set, what you will not do to the site, what the first 30 days contain); end the last FAQ with the booking CTA.

Dimension 7: Visual Hierarchy and Scannability

• Finding: A 30-second scan gets the story from the H2s “What an AI answer needs before it names your business”, “What each AI platform reads”, “The work ends on your WordPress website, not in a document”, which are benefit-led. Six H2s are label-led or split-phrase: “Be understood”, “Build confidence”, “Invite action”, “One system A complete view”, “Your questions Clear answers”, “Visibility is the start Business value is the goal”. Zero images on a 3,329-word page: no people, no work, no results shown. Paragraphs run 4 to 6 lines. Mobile layout renders cleanly (verified September 13).
• Severity: HIGH.
• Fix direction: convert the six label H2s to outcome statements; add one real image per major section (founder, a verified result, a rendered page).

Dimension 8: Friction Point Audit

• Finding: No form on the page; booking happens on /book-a-call/. Load is 1.0 second desktop, 0.5 phone. No pop-ups. Confirmation copy is explicit (“30 minutes with Phil. Clear next steps. No obligation.”). Navigation carries 50 links with three dropdowns, which is the one exit-ramp surface. Specificity is present where it matters (“30 minutes”, “since 1995”, “25 specialized”). No pricing anchor because no price is shown.
• Severity: LOW.
• Fix direction: no structural change; consider a lighter header on this page type.

Dimension 9: Search Intent Alignment (non-local page)

• Finding: The target “AI SEO services” is in the H1, title, URL, first sentence and the H2 “Business perspectives on AI SEO Services”. Service and FAQPage schema are present in the page’s structured data (verified from the stored graph). Ten supporting posts link up to this page through their endorsement anchors, and the page is linked from the header on every page. Intent is commercial-investigational and the page matches it: definition, mechanism, fit, proof, FAQ, booking.
• Scope boundary: this is a light on-page sanity check, not a ranking audit. No on-page ranking gap found. Off-page ranking work (the home cluster’s off-page corroboration) is a separate workstream and routes to the paired site audit, not attempted here.
• CRO and AI-citation overlap: the Phase 6 hero rewrites and the Phase 7 answer-first FAQ additions serve both conversion and AI citation.
• Severity: LOW.
• Fix direction: none on-page.

Dimension 10: VOC Sourcing Quality of Rewrites To be scored after Phase 6.

Dimension 11: Reading Level Compliance of Rewrites To be scored after Phase 6.

Dimension 12: Clutter and Navigation (rendered page)

• Finding: Rendered September 13, 2026 at 1366 by 768 and 390 by 844. Above-fold interference inventory: zero pop-ups, zero interstitials, zero sticky bars; one fixed scroll-to-top control (41 by 41 pixels, 0% of viewport). Nothing competes with the answer at load. Navigation: 50 links, three dropdowns, no dead ends; every dropdown item resolves to a live page. Minor friction only from menu density.
• Severity: LOW.
• Fix direction: none required.

Phase 4 Completion Gate: all four items YES. Ten page dimensions have findings citing Phase 1 evidence, every dimension carries a severity, Dimensions 10 and 11 carry placeholders.

---

Phase 5: Psychology Layer Overlay

Principle 1: Desire vs. Selling. PARTIAL. The hero builds desire (“discover and consider your business when they ask…”) and the section “What an AI answer needs before it names your business” builds understanding rather than pushing. The section “One system A complete view” and the capabilities sentence in the hero sell the vendor (“25 specialized analysis and implementation capabilities”) before the prospect has been shown a transformation they can picture. Opportunity: replace the capabilities sentence in the hero with one specific, dated result.

Principle 2: Scarcity and Urgency. NO. No real scarcity is stated (capacity, founder-led bandwidth, booking lead time). Manufactured urgency is forbidden; real founder capacity is a true constraint on a founder-led service. Opportunity: state the true constraint once, near the closing CTA, only if it is real (“Phil takes a limited number of new engagements each quarter” is acceptable only if true).

Principle 3: Loss Aversion. PARTIAL. VOC 11’s cost of inaction (“rank well on Google but remain invisible when someone asks ChatGPT”) is present in spirit in the hero, but the page never frames the loss in the buyer’s terms: the competitor gets named instead. Opportunity: one sentence in the fit section naming what happens when a competitor is the business the answer names.

Principle 4: Social Proof and Mimetic Desire. NO. No testimonial, no named customer, no similar-business identification on this page. The case-study paragraph names business types (“a special education law firm”, “a cell tower lease consultancy”, “a wellness spa”) which is the right shape, but it is one paragraph without an outcome number. Opportunity: bring one descriptor-named result with a figure onto this page beside the proof CTA.

Principle 5: Authority Signals. YES. “Since 1995” and “Founder led” are in the first screen; “Phil Belleville’s SEO experience dating to 1995” is in the hero. What is missing is any third-party authority (association, partner status, press).

Principle 6: Reciprocity. YES. “Book a free strategy call … 30 minutes with Phil. Clear next steps. No obligation.” Free value before the ask, stated plainly.

Principle 7: Commitment and Consistency. NO. There is no micro-commitment before the hard CTA: no self-check, no “which of these describes you” fit list the visitor can identify with before booking. Opportunity: convert “For businesses ready to be considered” into a three-item self-identification list directly above the closing CTA.

Phase 5 Completion Gate: all three items YES.

---

Phase 6: Copy Rewrites

Weak elements flagged in Phases 4 and 5: H1; hero subhead “Be part of the decision”; primary CTA “Book a free strategy call”; H2 “Your questions Clear answers”; H2 “For businesses ready to be considered”; H2 “Give the right buyers more reasons to consider you”. Six elements, three alternatives each, 18 alternatives.

Element 1: H1, currently “AI SEO Services”

Alternative 1: “AI SEO Services That Get Your Business Named, Not Just Ranked”

• VOC source: phrase 11, https://contently.com/2025/12/03/how-do-i-measure-visibility-in-llms-like-chatgpt-or-gemini/
• Mechanism: names the exact gap the prospect has already felt (ranking without being named) as the outcome; fires recognition before evaluation.
• Reading level check: read aloud, no stumble, 6th grade.

Alternative 2: “AI SEO Services for WordPress Businesses Tired of Paying for Reports”

• VOC source: phrase 2, https://au.trustpilot.com/review/bestseosolution.net
• Mechanism: mirrors the betrayal the burned buyer describes (“paying for empty promises”) and positions the page’s own promise, work installed on the site rather than reported.
• Reading level check: read aloud, no stumble, 7th grade.

Alternative 3: “AI SEO Services: Be the Business ChatGPT and Google Recommend”

• VOC source: phrase 15, https://www.onely.com/blog/best-ai-seo-agencies/
• Mechanism: states the outcome as the prospect pictures it, being the one named, in the platforms the prospect uses.
• Reading level check: read aloud, no stumble, 6th grade.

Element 2: Hero subhead, currently “Be part of the decision”

Alternative 1: “When a buyer asks an AI who to hire, be the name it gives.”

• VOC source: phrase 15, https://www.onely.com/blog/best-ai-seo-agencies/
• Mechanism: makes the abstract “decision” concrete: the buyer’s question and the answer’s name.
• Reading level check: read aloud, no stumble, 5th grade.

Alternative 2: “Ranking well and still not being named is the gap this closes.”

• VOC source: phrase 11, https://contently.com/2025/12/03/how-do-i-measure-visibility-in-llms-like-chatgpt-or-gemini/
• Mechanism: names the specific pain in the buyer’s words and promises the specific fix.
• Reading level check: read aloud, no stumble, 6th grade.

Alternative 3: “Built into the WordPress site you own, not written into a report.”

• VOC source: phrase 14, https://markets.financialcontent.com/clarkebroadcasting.mymotherlode/article/pressadvantage-2025-11-27-why-seo-partnerships-fail-seoptimize-shares-what-business-owners-really-want
• Mechanism: answers the buyer who “had no idea what their last agency was actually doing” with visible, installed work.
• Reading level check: read aloud, no stumble, 6th grade.

Element 3: Primary CTA, currently “Book a free strategy call”

Alternative 1: “Book 30 minutes with Phil”

• VOC source: phrase 4, https://au.trustpilot.com/review/bestseosolution.net
• Mechanism: the reviewer’s relief was working “directly with the owner”; the button names the person and the time cost.
• Reading level check: read aloud, no stumble, 4th grade.

Alternative 2: “Get a straight answer on fit, free”

• VOC source: phrase 3, https://au.trustpilot.com/review/bestseosolution.net
• Mechanism: promises what the burned buyer wanted (“money came before performance” reversed): an honest assessment first, no obligation.
• Reading level check: read aloud, no stumble, 5th grade.

Alternative 3: “See what your site needs, in one call”

• VOC source: phrase 14, https://markets.financialcontent.com/clarkebroadcasting.mymotherlode/article/pressadvantage-2025-11-27-why-seo-partnerships-fail-seoptimize-shares-what-business-owners-really-want
• Mechanism: value-led button copy that answers the “no idea what they were doing” wound with clarity as the deliverable of the call.
• Reading level check: read aloud, no stumble, 5th grade.

Element 4: H2, currently “Your questions Clear answers”

Alternative 1: “What buyers ask before they book, answered plainly”

• VOC source: phrase 7, https://community.shopify.com/t/why-does-seo-take-so-long/395499
• Mechanism: the skeptical buyer wants plain answers, not reassurance; the heading promises the form of answer they are looking for.
• Reading level check: read aloud, no stumble, 6th grade.

Alternative 2: “The questions a burned buyer asks first”

• VOC source: phrase 5, https://community.shopify.com/t/why-does-seo-take-so-long/395499
• Mechanism: names the reader’s history without judging it, which is the recognition trigger for a prospect on their second or third agency.
• Reading level check: read aloud, no stumble, 6th grade.

Alternative 3: “Guarantees, platforms, price, and what happens on your site”

• VOC source: phrase 9, https://www.teamblind.com/post/any-good-resources-recommendations-for-seo-optimization-nkimjcoh
• Mechanism: lists the four objections the bank surfaced, so the reader sees theirs before opening a single toggle.
• Reading level check: read aloud, no stumble, 6th grade.

Element 5: H2, currently “For businesses ready to be considered”

Alternative 1: “This fits you if your site ranks but never gets named”

• VOC source: phrase 11, https://contently.com/2025/12/03/how-do-i-measure-visibility-in-llms-like-chatgpt-or-gemini/
• Mechanism: a self-identification statement; the reader either recognizes it or self-ejects, which is what a fit section is for.
• Reading level check: read aloud, no stumble, 5th grade.

Alternative 2: “Who this is for, and who it is not”

• VOC source: phrase 13, https://misspepper.ai/ai-optimization/artificial-intelligence-optimization/compare-ai-seo-services-for-small-businesses-the-no-nonsense-guide-to-not-wasting-your-budget/
• Mechanism: the skeptic’s question (“basically just using ChatGPT to write blog posts”) is a fit question; a heading that promises both halves earns the read.
• Reading level check: read aloud, no stumble, 4th grade.

Alternative 3: “A real offer, real experience, and a WordPress site doing less than it could”

• VOC source: phrase 6, https://community.shopify.com/t/why-does-seo-take-so-long/395499
• Mechanism: lifts the page’s own three fit conditions into the heading so the reader tests themselves against them in one line.
• Reading level check: read aloud, no stumble, 6th grade.

Element 6: H2, currently “Give the right buyers more reasons to consider you”

Alternative 1: “Next step: 30 minutes with Phil, no obligation”

• VOC source: phrase 4, https://au.trustpilot.com/review/bestseosolution.net
• Mechanism: the closing heading states the action, the person and the risk removal in one line, mirroring the page’s own microcopy.
• Reading level check: read aloud, no stumble, 5th grade.

Alternative 2: “Find out what your site needs before you spend anything”

• VOC source: phrase 9, https://www.teamblind.com/post/any-good-resources-recommendations-for-seo-optimization-nkimjcoh
• Mechanism: addresses the price objection directly at the point of decision by sequencing diagnosis before spend.
• Reading level check: read aloud, no stumble, 6th grade.

Alternative 3: “Stop paying for reports. Start with a call about your site.”

• VOC source: phrase 2, https://au.trustpilot.com/review/bestseosolution.net
• Mechanism: the betrayal phrase inverted into the page’s differentiator, installed work over paper.
• Reading level check: read aloud, no stumble, 5th grade.

Phase 6 Completion Gate: all seven items YES. 18 alternatives, every one VOC-sourced, read aloud, no em dashes, no question-form headlines.

---

Phase 6A: Containment Gate

Ownership-bearing alternatives in the set: the three H1 alternatives. The title tag and meta description were not flagged weak and were not rewritten. The five H2 and CTA elements rewritten carry no pillar head and skip the gate by rule.

Command run for each: python3 pillar_gate.py --registry PILLAR_REGISTRY_aisearchengineoptimization.json --draft cro/draft_h1_N.json

Design note: every alternative keeps the exact Main “AI SEO Services” as its opening phrase and none contains “AI SEO agency” or “AI SEO company”, both fenced to the homepage. Zero discarded alternatives; each element still carries three.

Phase 6A Completion Gate: all four items YES.

---

Phase 7: Three-Tier Recommendations and Experiment Bank

7.1 Critical Fixes (3)

Critical Fix 1: no proof beside any CTA

• Problem: The hero CTA “Book a free strategy call” and the closing CTA “Book your free strategy call” have no trust element adjacent to them. The only proof on the page is the paragraph “Eight published case studies show Google ranking results … they do not establish AI citation or revenue gains”, mid-lower page, without a number (Phase 1.4, Dimension 5 CRITICAL).
• Solution: Beneath the hero microcopy, add one line drawn from the site’s own verified records, client by descriptor: “On one client site in August 2026 we installed 116 link changes across 77 pages and checked every one live.” Repeat one different descriptor-named result beside the closing CTA. Both figures already exist on the site’s published posts.
• Expected impact: HIGH. Proof adjacent to the ask is the single strongest lever on a page whose buyers arrive with VOC 5 and 8 (“zero positive results”, “zero accountability”). Direction: more bookings from the same traffic; magnitude cannot be estimated without a baseline.
• Evidence: Phase 1.4 (six NO rows), Phase 5 Principle 4 (NO), VOC 5, VOC 8.

Critical Fix 2: the H1 is a category label

• Problem: H1 “AI SEO Services” (Phase 1.1) fails the FIRE test (Dimension 1 HIGH, Phase 2 Attention MARGINAL).
• Solution: Replace with H1 Alternative 1, “AI SEO Services That Get Your Business Named, Not Just Ranked” (gate PASS). Keep the eyebrow. Keep the Main phrase leading so containment and ranking are unchanged.
• Expected impact: HIGH on attention retention at the door; the H1 now carries the outcome the supporting copy had to supply.
• Evidence: Phase 2.1 answers 1 and 3 (NO), Dimension 1 finding, Phase 6A gate record.

Critical Fix 3: a 3,329-word page with no CTA between screen one and the close

• Problem: CTA inventory (Phase 1.3) shows booking links in the header, hero and closing section only; the accountability section and the eight FAQs carry no CTA (Dimension 4 MEDIUM, compounded by page length).
• Solution: Add “Book 30 minutes with Phil” (CTA Alternative 1) as a button directly beneath the case-studies paragraph, and end the eighth FAQ answer (“How do we begin?”) with a linked “Book your free strategy call”. Same destination, /book-a-call/, no new path.
• Expected impact: MEDIUM to HIGH. A reader convinced at the proof section or the FAQ currently has to scroll to find the ask.
• Evidence: Phase 1.3, Dimension 4, Dimension 6.

7.2 High-Impact Changes (5)

1. Add three FAQs and a soft CTA on the last.
   • What: Beneath the existing eight, add “How is the price set?” (answer from the page’s own process: free call, written scope and price, month to month), “What will you never do to my site?” (answer: no purchased links, no unverifiable claims, every change verified live), and “What happens in the first 30 days?” (answer from the audit-first sequence). End the last FAQ with “Ready? Book your free strategy call.”
   • Why: VOC 9 (price), VOC 10 (harm), VOC 14 (not knowing what the agency does) are the three objections the bank surfaced that the FAQ does not name. Also serves AI citation: three more question-form blocks.
   • Implementation: three H3 and paragraph pairs appended inside the existing FAQ section; matching Question nodes added to the FAQPage schema.
   • Effort: LOW.
2. Convert the six label H2s to outcome statements.
   • What: “Be understood” to “The engines know exactly who you are”; “Build confidence” to “Buyers find reasons to trust you before they call”; “Invite action” to “The next step is obvious on every page”; “One system A complete view” to “One system reads your whole site the way an engine does”; “Your questions Clear answers” to H2 Alternative 1; “Visibility is the start Business value is the goal” to “Reported as what changed, not as one blended number”.
   • Why: Dimension 7 HIGH; label H2s fail the 30-second scan.
   • Implementation: six heading edits, no body change; run the three step labels past the card design because they also appear in the hero card.
   • Effort: LOW.
3. Put one real image in each major section.
   • What: A photograph of Phil at the top of “The work ends on your WordPress website”; a screenshot of a verified installed change (the internal-link workbook or a before/after schema render) in the accountability section; a rendered client page (descriptor-only, brand blurred) in “Pages built to the bar”.
   • Why: Dimension 7 finding, zero images on 3,329 words; Principle 4 and 5 (real people, real work).
   • Implementation: three images, each under 150 KB, each with descriptive alt text; featured image already set.
   • Effort: MEDIUM.
4. Move the trust strip up beside the hero CTA.
   • What: Place “Since 1995 | WordPress specialty | Founder led” as a single line directly beneath the CTA microcopy, ahead of the three-step card on phone.
   • Why: Dimension 3; the most specific elements in screen one currently sit below the CTA.
   • Implementation: one block reorder in the hero template; verify on phone that the CTA stays above the fold.
   • Effort: LOW.
5. Rewrite the fit section as a self-identification list with a micro-commitment.
   • What: Under H2 Alternative 1 (“This fits you if your site ranks but never gets named”), a three-item list the reader checks against: “You have a real offer and real experience.” “Your WordPress site does less than it could.” “Buyers compare you through AI answers as well as Google.” Then the closing CTA.
   • Why: Principle 7 (no micro-commitment); Dimension 2 (who-it-is-for arrives late).
   • Implementation: replace the fit paragraph with the list; keep the platform sentence beneath it.
   • Effort: LOW.

7.3 Test Hypotheses (5)

1. Hypothesis: an outcome-led H1 raises booking clicks from the hero.
   • Control: H1 “AI SEO Services”.
   • Variant: “AI SEO Services That Get Your Business Named, Not Just Ranked”.
   • Success metric: clicks on the hero CTA per session, 28 days, split by day; minimum 200 sessions per arm before reading.
   • Risk: the longer H1 wraps to three lines on phone; verify rendering first.
2. Hypothesis: a verified figure beside the hero CTA raises bookings more than the trust strip alone.
   • Control: microcopy “30 minutes with Phil. Clear next steps. No obligation.”
   • Variant: same microcopy plus the 116-changes line beneath it.
   • Success metric: /book-a-call/ page views originating from /ai-seo/, 28 days.
   • Risk: the figure is from one client; state it as one client, never as typical.
3. Hypothesis: a mid-page CTA after the proof section captures readers who currently leave.
   • Control: no CTA between hero and close.
   • Variant: “Book 30 minutes with Phil” button beneath the case-studies paragraph.
   • Success metric: scroll-depth-to-click ratio at the 60 to 70% mark, 28 days.
   • Risk: none material; same destination.
4. Hypothesis: naming the price process in a FAQ lowers bounce among price-sensitive readers.
   • Control: eight FAQs, no price FAQ.
   • Variant: “How is the price set?” FAQ added.
   • Success metric: exit rate from the FAQ section, 28 days, plus bookings.
   • Risk: readers expecting a number may still leave; the answer must state why there is no fixed figure.
5. Hypothesis: a self-identification fit list above the closing CTA raises closing-CTA clicks.
   • Control: fit paragraph as written.
   • Variant: three-item list under H2 Alternative 1.
   • Success metric: closing CTA clicks per session reaching the section, 28 days.
   • Risk: a reader who fails the list self-ejects; that is the intended effect and should be watched, not prevented.

7.4 Experiment Bank (15 tests, 5 categories, 3 each)

Category A: Hero section

• A1. H1 Alternative 1 vs Alternative 3 (outcome-led vs platform-named).
• A2. Subhead Alternative 1 (“When a buyer asks an AI who to hire, be the name it gives.”) vs current “Be part of the decision”.
• A3. Three-step card vs a single photograph of Phil in the right column; static image is the default recommendation.

Category B: Social proof

• B1. Verified-figure line beneath the hero CTA vs beneath the closing CTA vs both.
• B2. Case-study paragraph as prose vs as a three-row result strip (descriptor, figure, window).
• B3. Descriptor-named result (“a bathtub reglazing company in Los Angeles”) vs business-type-only (“a service business”).

Category C: CTA

• C1. “Book a free strategy call” vs “Book 30 minutes with Phil”.
• C2. One CTA path (booking) vs booking plus a visible email address for readers who will not book.
• C3. Sticky header CTA present vs header CTA hidden on this page.

Category D: Trust signals

• D1. Trust strip below the card (current) vs directly beneath the CTA microcopy.
• D2. “Since 1995” vs “31 years of SEO judgment” (the site’s ruling is “since 1995”; test only the placement, not the phrasing, unless the ruling changes).
• D3. Association or partner badge present in the hero vs absent (only if a badge applicable to this entity exists).

Category E: Friction, pricing and objection

• E1. Price-process FAQ present vs absent.
• E2. “What will you never do to my site?” FAQ present vs absent.
• E3. FAQ section above the closing CTA (current) vs FAQ placed directly after the proof section with the closing CTA beneath it.

Phase 7 Completion Gate: all seven items YES. 3 Critical, 5 High-Impact, 5 Hypotheses with control, variant and metric, 15 experiments across 5 categories with 3 each, every item naming a specific element on this page.

---

Phase 8: Composite Scoring

8.1 /ai-seo/ (the page under audit)

Verdict bucket: 72 to 95, “Decent foundation. High-impact fixes available.” No rebuild routing: composite is above 48 and only one dimension scored CRITICAL.

Page-dimension subtotal (Dimensions 1 to 9 and 12, the ten that score the page as it exists): 64 /100.

8.2 Competitor comparison, same ten page dimensions

Page: NP Digital, “AEO / GEO Agency Services”, https://npdigital.com/solutions/earned-media/ai-search-engine-optimization/, captured and rendered September 13, 2026. Dimensions 10 and 11 score rewrites and do not apply to a page that was not rewritten; the comparison is on the ten page dimensions, out of 100.

Result: /ai-seo/ scores 64 of 100 on the ten page dimensions against 45 of 100 for the competitor’s equivalent service page. The audited page leads on eight of the ten dimensions and trails on two: Trust and Proof (4 vs 6, the competitor shows one figure and this page shows none) and, marginally, nothing else; Search Intent is 8 vs 6 in this page’s favor. The single dimension where the competitor is stronger is exactly the one this audit scores CRITICAL, which is why Critical Fix 1 leads the list.

Phase 8 Completion Gate: all six items YES. Twelve numeric scores each with a note, composite summed correctly (4+7+5+7+4+7+5+8+8+9+9+9 = 82), verdict named, Dimensions 10 and 11 scored from the actual Phase 6 state.

---

Phase 9: Pre-Delivery Self-Check (17 items)

---

Phase 10: Cumulative Execution Audit

Every phase PASS. The deliverable is ready.