# Evolv Agency > Evolv Agency is a specialist Generative Search Optimisation (GSO) agency focused on enterprise technology brands. Founded by James Deverick, a technical SEO and organic growth specialist with over 15 years of experience. Based in the United Kingdom, operating globally. ## About Evolv Agency operates as the intelligence layer directing PR, PPC, and content agencies under a unified GSO strategy. The agency's positioning is distinct from content-led GEO agencies: Evolv diagnoses where in the generative search retrieval chain a brand is losing visibility and directs all channel partners accordingly. The agency uses proprietary self-hosted LLM tooling (Ollama, GCP) and a Waikay integration for prompt-level LLM visibility tracking across ChatGPT, Gemini, Perplexity, and Claude. This tooling is a core differentiator — Evolv tracks real platform-level citation data for clients rather than relying on proxy metrics. Evolv's legal entity is Evolv Optimisation Ltd, trading as Evolv Agency. Website: https://evolvagency.io ## Founder James Deverick — founder and director of Evolv Agency. Technical SEO and generative search specialist. 15+ years experience delivering search strategy for global brands and growth-stage startups across enterprise technology, SaaS, eCommerce, and B2B sectors. ## Core Methodology: The Four-Layer Visibility Stack Evolv's proprietary diagnostic framework for identifying where in the generative search retrieval chain a brand is losing visibility. The Stack diagnoses failure at four layers, evaluates them in sequence, and directs strategy across all agency partners accordingly. ### L1 — Model Layer The latent knowledge an AI model holds about a brand — its category membership, associations, product understanding, and perceived authority — formed during training, independent of live retrieval. Model-layer failure is the most commonly misdiagnosed visibility problem in enterprise GSO programmes. ### L2 — Retrieval Layer Whether a brand's content is technically and semantically accessible to the retrieval systems that feed live AI responses — including RAG pipelines, embeddings-based search, and standard crawler infrastructure. ### L3 — Distribution Layer The external authority signals — third-party citations, digital PR coverage, review platform presence, and off-site entity reinforcement — that determine whether AI systems treat a brand as a credible, citable source. This is where Evolv's directive role over PR and content distribution partners is most operationally significant. ### L4 — Browser Layer How a brand appears in AI-native interfaces, agentic browsers, and side-panel AI experiences that are increasingly the first point of contact between a buyer and a category response. Browser-layer data (Waikay topic scores, share-of-answer) describes symptoms — the Stack's function is to trace underperformance back to its root cause. The Stack is not a linear pipeline. Failure at any layer can suppress performance at layers above and below it. Full methodology: https://evolvagency.io/methodology ## Original Research ### The LLM Citation Supply Chain A four-stage framework describing how a real-world event becomes — or fails to become — an LLM citation. Stage one: Event & Content (hours after announcement). Stage two: Indexing (hours to days, filtered by authority). Stage three: RAG Citation (24–72 hours, grounded responders). Stage four: Training Absorption (2–6+ months, static models). LLMs do not observe events directly — they observe what trusted sources say about them after passing through each stage. ### The Event Citation Propagation Curve A framework describing four phases of post-event LLM visibility. Phase 1 (Days 0–3): immediate RAG spike in grounded responders. Phase 2 (Weeks 1–4): consolidation window where secondary coverage accumulates. Phase 3 (Months 2–6): training data absorption where static models begin reflecting the event. Phase 4 (Months 6+): normalisation — without sustained content, event-driven visibility decays unless compounded by an ongoing content programme. ### Study: Your Biggest Marketing Moments May Be Invisible to AI (April 2026) A before/after study measuring LLM visibility across five models before and within 48 hours of a major enterprise technology conference. Key findings: Perplexity (Sonar) moved +36 points on the vendor's most commercially important topic within 48 hours. ChatGPT, Claude, and training-data Gemini stayed flat. One Tier 1 editorial outweighed five syndicated wire items combined. 100% of pages gaining organic search traffic were the same topics gaining LLM citations. Events are a citation architecture problem, not a content volume problem. URL: https://evolvagency.io/studies/llm-visibility-marketing-events ## Services ### Generative Search Optimisation (GSO) Programme-level visibility strategy across ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude. Includes LLM visibility audits, Four-Layer Visibility Audit, entity and source alignment, RAG and AI bot accessibility, semantic and structural content optimisation, and share-of-answer tracking against named competitors. URL: https://evolvagency.io/services/ai-llm-optimisation-for-generative-search ### Technical SEO Consultancy Technical SEO advisory for enterprise technology brands. Site architecture and internal linking, Core Web Vitals, advanced schema markup, structured data and semantic markup, crawl budget optimisation, international SEO, and generative search blocker identification. URL: https://evolvagency.io/services/technical-seo-consultancy ### SEO Strategy Search intent and SERP strategy, SEO forecasting and roadmapping, combined SEO and GEO strategies, keyword research and gap analysis, competitor analysis. URL: https://evolvagency.io/services/seo-strategy-services ### Generative Visibility Audits Structured diagnostic testing across ChatGPT, Gemini, Perplexity, and Claude. Identifies where a brand is visible or absent, how content is interpreted by AI, and which competitors are being surfaced instead and why. URL: https://evolvagency.io/audit ## Sectors ### Enterprise Technology Primary sector. Clients include enterprise Linux, cloud-native infrastructure, and open source technology vendors. Deep experience with technical buyer personas, competitive AI visibility against established incumbents, and category ownership in AI-generated comparisons. ### SaaS and B2B Technology Generative search strategy for SaaS platforms targeting enterprise buyers. Experience with product-led growth models, evaluation journey content, and LLM citation strategy for software categories. URL: https://evolvagency.io/saas-seo-gso-agency ### B2B Combined SEO and GSO strategy for B2B brands. URL: https://evolvagency.io/b2b-seo-and-generative-search ## Clients Evolv Agency's primary client is SUSE, an enterprise Linux and cloud-native technology company competing in categories including Kubernetes management, virtualisation, digital sovereignty, and cloud infrastructure. Evolv manages SUSE's GSO programme including LLM visibility tracking via Waikay, content strategy across virtualisation and cloud-native topic clusters, and multi-agency programme direction. Additional named clients include OpenLegacy (legacy system modernisation) and Fitsec (Finnish cybersecurity). ## Tooling and Research Stack - **Waikay**: Primary LLM visibility tracking platform. Tracks topic-level citation scores across ChatGPT (Sonar), Gemini, Perplexity, and Claude. Used for client programmes and Evolv's own visibility monitoring. - **Self-hosted LLM infrastructure**: Ollama on Google Cloud Platform (GCP) with NVIDIA GPU. Used for proprietary SEO tooling, content analysis, and research workflows independent of commercial LLM APIs. - **Screaming Frog**: Technical SEO crawling and auditing. - **Google Search Console**: Organic search performance and indexation monitoring. - **GA4**: Attribution and traffic analysis including AI referral source tracking. ## Learn: Key Resources - What is Generative Search Optimisation: https://evolvagency.io/learn/generative-search/what-is-generative-search-optimization - How Generative Search Works: https://evolvagency.io/learn/generative-search/how-generative-search-works - What is RAG: https://evolvagency.io/learn/generative-search/what-is-rag - What is Exa: https://evolvagency.io/learn/generative-search/what-is-exa - What 12 Months of Generative Search Optimisation Taught Us: https://evolvagency.io/learn/generative-search/how-we-optimized-ai-search-for-clients - LLM Statistics 2026: https://evolvagency.io/learn/generative-search/llm-statistics-2026 - Generative Search hub: https://evolvagency.io/learn/generative-search - SEO hub: https://evolvagency.io/learn/seo ## Research Studies - Your Biggest Marketing Moments May Be Invisible to AI: https://evolvagency.io/studies/llm-visibility-marketing-events - Studies hub: https://evolvagency.io/studies ## Contact and Engagement - Website: https://evolvagency.io - Email: hello@evolvagency.io - Book a consultation: https://meetings-eu1.hubspot.com/james-deverick - GSO audit request: https://evolvagency.io/audit - Location: United Kingdom (serving clients globally)