Before we run the audit, we need to make sure we're asking the right questions about the right competitors to the right buyers. This document presents what we've learned about Retina Media's market — your job is to tell us what we got right, what we got wrong, and what we missed.
Before we measure citation visibility in the boutique GEO consultancy space, these three signals tell us whether AI crawlers can reach, read, and trust retina.media. They anchor everything that follows — and for a firm that sells GEO, they double as a credibility check on your own house.
AI search is rewriting how B2B technology buyers assemble vendor shortlists — they now ask ChatGPT, Perplexity, Claude, and Google AI Overviews who to trust before a single sales conversation. As a boutique Generative Engine Optimization strategy consultancy, Retina Media sells visibility in exactly this arena, which raises the stakes twice over: the firm competes for citations in an emerging category where early movers compound their advantage, and its own site is the reference implementation prospects and AI models judge it by. Establishing durable AI visibility now — for Retina and for the clients it serves — locks in trust that becomes self-reinforcing as platforms learn which domains to cite.
This Foundation Review presents three things we need to validate together before the audit runs: the competitive landscape that shapes how we build head-to-head queries, the buyer personas that determine the search intent patterns we simulate, and the Layer 1 technical baseline that determines whether AI platforms can access and extract retina.media's content at all. Each section exists so you can confirm or correct an input before it drives hundreds of buyer queries — this is the moment to catch anything we read wrong from the outside.
The validation call is a working session with real stakes. It resolves two kinds of decisions: input validation (are the right buyers, competitors, features, and pain points in the right tiers?) and engineering triage (which technical fixes can your team start before results come back?). Your answers set the architecture of the entire audit — the specific items to confirm and the tasks to start are in the TL;DR below and the Pre-Call Checklist at the end.
This is a validation document, not a report card. Everything here is our outside-in read of the boutique GEO consultancy market — your job is to confirm, correct, or add.
What This Is This Foundation Review captures what we've assembled about Retina Media's position in the boutique GEO strategy consultancy category: who your buyers are, who you're compared against, what you do well, and what frustrations drive buyers to look for a firm like yours. We built it from your site and public signals before running the audit so that every query we generate is grounded in your actual market.
What We Need From You Read each section and react. Where a purple box asks a question, it's flagging a genuine uncertainty whose answer changes how we build the audit. Tell us what we got right, what we got wrong, and — most valuable — what we missed. Corrections here are cheap; corrections after the audit runs are not.
How Confidence Badges Work Every entity carries a confidence badge. High means we sourced it directly from your site or a strong public signal. Medium means we inferred it from category patterns and want your confirmation. Medium-confidence items are exactly where your validation adds the most value.
The baseline identity that anchors every query. If we've mislabeled the category or segment, the whole query set drifts.
→ We modeled Retina as a startup, but your clients reportedly range from startups to enterprise — which client tier should the audit's buyer set represent? If we're modeling enterprise buyers, personas move up in seniority (VP/C-suite, longer committees, risk/procurement language); if startups, we weight founder-led, budget-conscious, speed-to-value vocabulary. This one answer cascades into personas, competitor relevance, and query phrasing.
5 personas: 2 decision-makers, 1 evaluator, 2 influencers. These personas drive the query set — each one searches differently, and the audit simulates their distinct intent patterns.
Critical Review Area All five personas are inferred, not observed. Retina is a boutique consultancy with no G2 or Capterra footprint, so we derived these buyers from standard B2B SaaS marketing buying-committee patterns rather than mining review data. Every persona below is Medium confidence — this is the section where your correction is worth the most, especially on who actually signs.
Data Sourcing Note Names, roles, seniority, veto power, and technical level are proposed from category patterns (source: LLM inference). Buying jobs and query focus areas are synthesized from those attributes plus your feature and pain-point set. Nothing here is scraped from a review platform — treat all of it as a hypothesis to confirm.
→ Two personas carry veto power — Lena and founder Aaron Feld. In a boutique deal, do both truly sign, or is one the recommender? If both, we split validation-stage queries across two approval lenses (marketing-outcome and founder-vision); if one, we collapse them and stop diluting the set.
→ We rated Marcus a high-influence evaluator without veto — does demand gen actually help shortlist GEO vendors, or only execute after marketing leadership picks? If he shortlists, we add mid-funnel "agency vs. platform for pipeline" queries in his voice; if not, we drop his cluster to save query budget for the true deciders.
→ On a content-led purchase, the content head often effectively drives vendor choice despite carrying no veto. Does Priya's judgment make or break the deal? If yes, we promote her cluster and add content-quality evaluation queries at her altitude rather than treating her as background influence.
→ Is Daniel the credibility gatekeeper who kills "GEO is just SEO hype" vendors, or a downstream implementer who inherits the decision? If he gates, his skepticism queries move earlier in the funnel and we weight methodology-defense language across the set.
→ In your deals, is the Founder/CEO the one who signs, or is this a marketing-led purchase where the CMO owns the decision? If founder-led, we shift query vocabulary toward executive and revenue framing; if marketing-led, we weight practitioner and demand-gen terms instead.
Who's Missing? These roles sometimes appear in boutique GEO deals — do they show up in yours? Head of RevOps / Marketing Ops (if pipeline reporting and attribution is a distinct buying conversation from creative and content). Head of Product Marketing (if category and positioning ownership sits with PMM rather than the CMO — directly relevant to your positioning offering). Fractional CMO / Chief of Staff (common in the startup clients you serve, and often the actual signer). Each would warrant its own query cluster. Who else shows up in your deals?
9 competitors: 5 primary, 4 secondary. Tier assignments determine which head-to-head matchups the audit tests.
Why Tiers Matter Primary tier means the audit builds direct-comparison queries against that vendor — roughly 6–8 head-to-head prompts each, like "best GEO agency for B2B SaaS" and "Retina Media vs. [competitor]." With 5 primaries, that's on the order of 30–40 head-to-head queries. Your set deliberately mixes done-for-you agencies (Omniscient Digital, Animalz, Foundation, Minuttia) with a self-serve platform (Profound) because you explicitly position against "generic GEO platforms" — but that mix is the biggest open question in this document. We're least certain about Minuttia's primary placement (medium confidence), and about whether Profound — a tool, not an agency — belongs in the head-to-head set at all.
→ Three things to settle: (1) Profound is a self-serve platform, not an agency — in real deals, are you compared against tools like Profound, or mostly against other done-for-you shops? If the former, we keep it primary and generate "agency vs. platform" head-to-heads; if not, it moves to secondary and ~6–8 queries reallocate. (2) Minuttia is our lowest-confidence primary — does it actually surface in your deals, or is it a lookalike we over-weighted? (3) Anyone we're missing (a specific agency that keeps beating you), or any listed vendor that's simply irrelevant to how you sell?
12 buyer-level capabilities mapped. These determine which capability queries the audit tests and where it probes for competitive vulnerability.
Show me exactly where my brand does and doesn't appear when buyers ask AI for a vendor shortlist, by persona and platform.
Run the real questions my buyers would type into ChatGPT and Perplexity across every persona and buying stage, not a handful of generic prompts.
Tell me which competitors AI recommends instead of me and how often they get cited over us.
Build content and structured assets that large language models can actually ingest, cite, and represent accurately.
Make my category and value proposition so clear that AI describes us the way we describe ourselves.
Don't just hand me a report — deploy the content and improvements for me every month so my visibility compounds.
Get my founders and execs publishing points of view that sound human and earn trust with AI models and readers.
Crawl my site and tell me what's blocking AI crawlers from reading and citing my pages.
Cover every AI engine my buyers actually use, not just one or two.
Connect AI visibility to actual leads, pipeline, and revenue — not just a rising visibility score.
Give me an always-on dashboard that tracks my AI citations in real time between engagements.
Earn the third-party mentions and high-authority links that make AI models treat my brand as credible.
Which Strengths Do We Emphasize? The audit tests all 12 capabilities, but competitive-differentiation queries will emphasize 3. Six are rated Strong — which of these best represents where Retina wins deals?
• AI Visibility Benchmarking & Audit
• Buyer-Intent Query Simulation
• Competitive Share-of-Voice Analysis
• LLM-Ready Content Production & Content Systems
• Positioning & Messaging Strategy
• Managed Execution & Ongoing Content Deployment
→ Two calibration questions: (1) We rated Real-Time Monitoring and Digital PR / Off-Site Authority as your weak spots versus the platforms (Profound, Scrunch, Goodie) and PR-led agencies — is that fair, or do you now offer a live dashboard or attribution you'd want us to test as a strength? These ratings decide where the audit probes for vulnerability; if they're wrong we'll under-test capabilities you actually win on. (2) Do AI Visibility Benchmarking and Buyer-Intent Query Simulation read as one capability to your buyers, or two distinct ones worth separate query clusters?
10 pain points: 5 high, 5 medium severity. The buyer language here is how the audit will actually phrase queries — in the buyer's words, not marketing copy.
→ Three checks: (1) Is the severity right — should "zero-click traffic erosion" and "can't prove GEO ROI" both stay High, or does one actually dominate the sales conversation? Whichever is truly High gets tested first. (2) Does the buyer language sound like your actual buyers, or too polished? We query in these exact words. (3) Missing pains we'd suggest for this category: fear of AI fabricating facts about the brand (defensive GEO), losing a specific high-intent branded query to a competitor's comparison page, and board pressure to "have an AI strategy" without a clear internal owner. Do any of these come up in your deals?
What our crawl found about whether AI engines can reach, read, and extract retina.media. These are technical hand-offs for engineering — not content strategy, which waits for audit results.
Actionable Now — Engineering No critical crawl blockers: robots.txt is present and explicitly allows every major AI crawler (GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot, Google-Extended, Googlebot). The work is structural. Engineering should verify three things before the call: (1) that the homepage isn't client-side-render dependent — our fetch returned only the hero line; (2) whether JSON-LD schema is actually present on posts, /cited, and /geo (our method can't see it); and (3) trim the 79 thin /blog/tag/* pages from the sitemap. The one high-severity item — stale, pre-repositioning proof pages — is a content refresh owned by your content team.
What we found: The eight portfolio/case-study pages carry sitemap lastmod dates ranging from 2018 to mid-2025. Seven of eight (Zenreach 2025-04, IDVerse 2025-04, SoFi 2024-05, HP 2021-06, Coca-Cola/AMC 2019-09, Southern Company 2020-04, University of Georgia 2018-08) are older than 365 days and score 0.0 on freshness. None carries a visible on-page date, and all predate the firm's GEO repositioning — they document legacy creative-agency work, not GEO engagements.
Why it matters: Case studies and portfolio proof are among the assets buyers and AI assistants reach for when validating a vendor. Content freshness is a strong AI-citation signal: Ahrefs finds AI-cited content is ~25.7% fresher on average than non-cited content (Ahrefs, August 2025), and ConvertMate found 76.4% of ChatGPT's most-cited pages were updated within 30 days (ConvertMate, ChatGPT-scoped). Legacy pages dated 2018–2022 are effectively invisible to freshness-weighted citation and, worse, misrepresent a GEO consultancy as an old-line ad shop.
Recommended fix: Add visible published/updated dates to every proof page, refresh the roster to lead with GEO-era engagements, and either retire or clearly archive pre-repositioning creative work. At minimum, re-stamp lastmod when content is genuinely reviewed.
What we found: The /portfolio and /university-of-georgia pages render the site name "Retina Media" as the H1 rather than a descriptive page-topic heading, and the /geo page renders two H1s ("Generative Engine Optimization" and a headline). On /university-of-georgia the true topic sits in an H2 with the section detail nested under H3, inverting the hierarchy.
Why it matters: The H1 is a primary signal LLMs and search crawlers use to identify what a page is about and to label an extractable passage. A generic site-name H1 wastes that signal and makes the page harder to classify and cite; duplicate H1s dilute topical focus.
Recommended fix: Set a single, descriptive, topic-bearing H1 on every commercial page (e.g. "University of Georgia — Campus Life" rather than "Retina Media"), and demote the hero headline on /geo to an H2 so there is exactly one H1 per page.
What we found: Several high-traffic entry pages are structurally thin: the /portfolio hub (content depth 0.2) is a bare image-link grid with a single service sentence, the homepage (0.2) is a one-line hero splash, /contact (0.0) is a form, and /press-awards (0.2) is a satirical list. These pages contain few or no self-contained passages an LLM could extract.
Why it matters: The homepage and portfolio are the most likely landing points for an AI crawler exploring the domain. When they contain almost no extractable prose, the crawler gets little to represent the brand with, and internal link equity flows to pages that can't themselves be cited.
Recommended fix: Add a substantive, citable introduction to the /portfolio hub (what Retina does, for whom, with what outcomes) and expand the homepage beyond the hero line with a short, extractable summary of the GEO offering. Contact/press pages can remain thin — they are not citation targets.
What we found: sitemap.xml (134 URLs) includes 79 /blog/tag/* taxonomy pages and one /blog/category/* page alongside ~50 real content URLs. robots.txt disallows the query-parameter tag form (/*?tag=*) but the path-based /blog/tag/ pages are crawlable and explicitly submitted in the sitemap.
Why it matters: Thin, near-duplicate taxonomy pages submitted in the sitemap spread crawl attention across low-value URLs and can surface as thin results. On a boutique site with ~50 substantive pages, 80 thin taxonomy URLs is a majority of the submitted set.
Recommended fix: Exclude /blog/tag/* and /blog/category/* from sitemap.xml (or noindex them) so the sitemap advertises only substantive, citable content. Squarespace allows this via page settings or a curated sitemap.
The following items could not be assessed through our analysis method (rendered markdown). We recommend your engineering team verify these manually before the validation call.
What to check: Fetching the homepage returned only the hero line ("When buyers ask AI who to trust, we make sure it's you.") and a single CTA link — far less content than the interior pages, which rendered fully. This may simply be a minimal splash design, but the sparse render is also consistent with content that only appears after client-side JavaScript execution. Our fetch method cannot distinguish the two.
Recommended action: Load the homepage with JavaScript disabled (or view page source / use Google's URL Inspection "View crawled page") to confirm the intended content is present in the initial HTML. If it is not, move critical copy into server-rendered markup.
What to check: Our analysis reads rendered markdown, so inline JSON-LD schema blocks are not visible to us. We could not confirm whether Article schema is present on the 26 blog/white-paper posts, FAQ schema on the /cited and /geo FAQ sections, Book/Product schema on /cited, or Organization/Person schema sitewide.
Recommended action: Run the key page types through Google's Rich Results Test / Schema Markup Validator. Confirm (or add) Article schema with datePublished/dateModified and author on posts, FAQPage schema where FAQs exist, and Book schema on /cited.
What to check: Rendered markdown does not expose <meta name="description">, Open Graph, or Twitter Card tags, so we could not evaluate whether pages have unique, descriptive meta descriptions and complete social-preview metadata.
Recommended action: Spot-check pages with a social-preview tool or view-source, confirming each key page has a unique meta description and complete OG image/title/description tags.
Partial Assessment Schema coverage could not be scored on any of the 41 pages (our method reads rendered markdown), and freshness is unscored on 5 pages (all 4 structural pages and 1 undated product page). Treat schema and those freshness values as "verify manually," not as passing — the Manual Verification Checklist above covers exactly these.
Why Now GEO is a timing play, and the window is open:
• AI search adoption is accelerating — buyer discovery patterns are shifting quarter over quarter.
• Early citations compound: domains AI platforms learn to trust now get cited more often as that trust reinforces itself.
• Competitors who establish GEO visibility first create a structural disadvantage for late movers.
• The boutique GEO consultancy space is still early-innings — acting now means competing against inaction, not against entrenched strategies.
The full audit will measure Retina's citation visibility across the buyer queries in this document — from "best GEO agency for B2B SaaS" and "Retina Media vs. Profound" to "is GEO real or repackaged SEO" and "prove GEO ROI to leadership." You'll see exactly which queries return answers that include your competitors but not Retina, on which engines, and what it would take to appear in them. Fixing the Layer 1 technical items now — the stale proof pages, the thin entry pages, the sitemap — improves your baseline before we even take the measurement, so the audit reflects the site at its best.
45–60 minutes. We walk through this document together, confirm the inputs, and resolve the open questions — competitor tiers, the true buyer, and the feature strengths that drive the query set.
We generate buyer queries across the validated personas, competitors, features, and pain points, then run them across the selected AI platforms (ChatGPT, Perplexity, Claude, Google AI Overviews).
Visibility analysis, competitive positioning, and a prioritized three-layer action plan — the point at which content recommendations are prioritized by which gaps actually cost you citations.
Start Now — Engineering Three Layer 1 fixes don't depend on the rest of the audit and will improve your baseline visibility before we even measure it: (1) confirm the homepage isn't client-side-render dependent — load it with JavaScript disabled and, if the hero copy is injected client-side, move it into server-rendered markup; (2) trim the sitemap by excluding the 79 /blog/tag/* pages and 1 category page; (3) verify JSON-LD schema on posts, /cited, and /geo via Google's Rich Results Test and add Article/FAQ/Book schema where missing. robots.txt already allows every major AI crawler, so no crawler-access remediation is needed — but re-confirm it hasn't changed before the call.
Two jobs before we meet. The questions on the left require your judgment — no one knows your business better than you. The engineering tasks on the right don't require the call at all.