Our complete GEO scoring framework, service delivery methodology, expertise credentials and South African market benchmarks — documented for full transparency.
AI Score was founded by a Project Manager, Agile Coach and Trainer with two years delivering AI training programmes inside corporate South Africa. This practical AI implementation experience — building structured AI adoption frameworks for large South African organisations — forms the bedrock of every GEO methodology we apply.
Our approach draws on direct experience of how AI tools are actually adopted inside South African enterprises: which signals build trust with AI systems, which content structures produce consistent citations, and which technical foundations are non-negotiable for AI visibility. This is not theory adapted from overseas research — it is a methodology developed through testing in the South African context.
Every engagement is delivered by the founder personally. No offshore teams. No junior analysts. The person who built the methodology delivers it — ensuring consistent, rigorous execution across every client.
Certified project management methodology applied to GEO service delivery. Every client engagement is managed as a structured project — defined milestones, clear deliverables, transparent progress reporting. No vague "we're working on it."
Agile methodology applied to iterative GEO optimisation. GEO is not a one-time project — it is a continuous improvement cycle. Our monthly retainer is structured as an Agile sprint: audit, implement, measure, iterate. Each sprint builds on the last.
Two years building and delivering structured AI training programmes inside major South African organisations. Developed frameworks for AI adoption that produce measurable business outcomes — not surface-level awareness. This insider knowledge of how AI is evaluated and deployed at enterprise level informs every GEO recommendation we make.
Founder of a dedicated AI and Agile training practice serving South African organisations. Programmes designed to translate AI capability into practical, measurable business outcomes. The same principle applies to AI Score: AI tools only produce results when applied with the right strategic framework.
A weighted composite of six dimensions. Each dimension reflects a different aspect of how AI engines discover, evaluate and cite a business.
Measures how extractable and answer-ready your content is for AI engines. Evaluated across: answer-first content structure (are questions answered in the first sentence?), FAQ hub depth (20–30 specific questions with substantive answers), statistical density (are there quantifiable, citable facts?), self-containment (can each content block be cited without surrounding context?), and uniqueness (is the content distinctive enough to cite specifically, not just as a generic category example?).
AI engines build entity understanding from brand mentions across high-authority platforms. Scored across: YouTube presence (AI engines frequently cite video content for how-to queries), Reddit engagement (Perplexity heavily weights authentic Reddit discussions), LinkedIn Company Page (critical for B2B Bing Copilot and Gemini citations), Wikipedia/Wikidata presence (enables Knowledge Graph integration), and NAP (Name, Address, Phone) consistency across directories. South African businesses without consistent entity presence score significantly lower on this dimension.
Google's Experience, Expertise, Authoritativeness and Trustworthiness framework is directly applied by AI citation algorithms. Evaluated across: Experience (first-person case studies, client outcomes, original data), Expertise (author bylines with credentials, About page completeness, professional certifications displayed), Authoritativeness (external citations, industry mentions, professional associations), and Trustworthiness (HTTPS, POPIA compliance statement, privacy policy, physical address, contact information). Service-based businesses in South Africa frequently fail this dimension by having no visible team credentials or About page.
AI engines must be able to crawl and render your site before they can cite it. Evaluated across: crawlability (internal linking structure, no dead-end pages), indexability (canonical tags, no noindex directives blocking key content), Core Web Vitals (LCP under 2.5s, CLS under 0.1), security headers (HTTPS, HSTS, CSP), mobile responsiveness, and server-side rendering. South African sites frequently fail on crawlability due to single-page architectures with no internal linking.
Schema.org structured data is the machine-readable layer that allows AI engines to understand entity relationships, service offerings and organisational context. The most critical schema types for South African SMEs: Organization (with sameAs links to social profiles), LocalBusiness (with geographic coordinates and service areas), Service (with pricing and descriptions), FAQPage (high-citation-value for conversational AI queries), and BreadcrumbList (navigation context for crawlers). Zero schema is the single most common critical failure we find in SA sites — directly responsible for low citation rates in Gemini and Bing Copilot.
Measures explicit AI-readiness signals. Evaluated across: AI crawler directives in robots.txt (GPTBot, ClaudeBot, PerplexityBot, Google-Extended — explicitly allowing or managing 14 major AI crawlers), llms.txt file (emerging standard for AI-friendly content declaration), Open Graph tags (og:title, og:description, og:image, og:url — required for social and AI preview rendering), canonical tags (prevents duplicate content confusion across AI indexing), content date metadata (published_date and modified_date — AI engines prioritise recency), and Bing Webmaster Tools verification (critical for Bing Copilot indexation speed).
Average AI Score across South African SMEs by sector, based on audits conducted by AI Score using Seomator GEO and our internal audit engine.
Scores represent a composite average. Individual business scores vary significantly based on website age, existing content depth and technical infrastructure. Data current as of April 2026.
Every Foundation Launch follows a structured 6-phase methodology. Each phase has defined inputs, outputs and success criteria.
Pre-sale · Free AI Visibility Check
Input: Business URL. Output: AI Score report (6-dimension breakdown, competitor comparison, top 3 quick wins). Delivery: Automated, within 60 seconds. Purpose: Demonstrates the gap clearly — data sells the service without any pitching required. Approximately 40% of free audit recipients proceed to a paid engagement.
Week 1 · Brand Voice Capture
Input: Client brand documentation, website, existing content. Session: Structured interview extracting distinctive brand voice, specific service claims, target query types, and competitive differentiation. Output: Brand Voice Document — the brief that governs all content produced in Phases 3–5. Why it matters: AI produces the statistical average of the internet. Without this layer, every business in a sector sounds identical to AI — and identical businesses receive random, inconsistent citations.
Weeks 1–2 · AI Visibility Audit
Scope: 5 AI engines, 3 competitor benchmarks. Tools: Seomator GEO, AI Score internal engine. Output: 15-page audit report — current score by dimension, competitor gaps, prioritised action plan with effort vs impact matrix. Delivery: PDF report + 30-minute debrief call walking through findings.
Weeks 2–4 · GEO Infrastructure Build
Deliverables: Organization + LocalBusiness + Service + FAQPage schema markup, AI crawler directives (14 crawlers), llms.txt, robots.txt, canonical tags, Open Graph + Twitter card tags, Bing Webmaster Tools verification, FAQ hub (20–30 questions with 150-word answers), answer-first content restructuring of key pages, internal linking architecture. Implementation: WordPress sites via WPCode plugin (no developer needed). Non-WordPress sites receive an Implementation Pack — ready-to-paste code with step-by-step instructions.
Month 2 · Brand AI Stack Build
Deliverables: Pillar-cluster content strategy, prompt library (calibrated to brand voice document from Phase 1), content calendar (12 months), topic cluster map, 4 citation-optimised articles published (1,500+ words each, answer-first structure, FAQ schema, statistical density targets met). Why brand voice matters here: AI engines cite specific, distinctive sources. Content that sounds generic receives random citation — content that sounds distinctively like your business receives consistent citation.
Ongoing · Monthly Retainer — R7,500/month
Week 1: 4 articles generated (brand-voice calibrated), AI humanised, published or exported. Week 4: Monthly AI Score report — score by dimension, citation frequency across 5 engines, 3-competitor benchmark, trend trajectory. Also included: One 45-minute strategy call. Quarterly full re-audit. Month-to-month — no lock-in. We earn continued engagement through results, not contracts.
AI Score uses independent third-party audit tools to measure and validate every score improvement. We do not rely solely on our own audit engine — every material change is cross-validated with external tools before being reported to clients.
Independent GEO scoring across citability, technical SEO, schema, E-E-A-T, brand authority and platform readiness. Used as the primary external benchmark for before/after score validation.
Proprietary audit tool built for SA market benchmarks. Crawls the homepage, checks all 6 dimensions, cross-references against sector benchmarks, and generates the free instant AI Score report.
Paid Perplexity subscription used to manually verify citation appearance — simulating exactly what a customer sees when asking a relevant query. The only reliable way to confirm Perplexity citation.
AI referral traffic tracking via UTM parameters. GSC validates crawl coverage. Bing Webmaster Tools used for Copilot indexation and IndexNow protocol.
Start with a free instant audit — no commitment, results in 60 seconds.