GEO: Generative Engine Optimisation for Trades

Search has quietly changed. When a homeowner needs a plumber at 9 pm on a Sunday, they are increasingly likely to type a question into ChatGPT or Perplexity rather than scroll through Google's blue links. What they get back is not a list of websites — it is a synthesised paragraph that names two or three businesses and explains why they are trustworthy.

That paragraph is the new first page of Google. And if your business is not in it, you do not exist for that customer.

This article explains what Generative Engine Optimisation (GEO) is, how it differs from the SEO you already know, and what you need to do — specifically as a trades business — to be cited by AI search tools.

What Is a Generative Engine?

A generative engine is a search tool that uses a large language model (LLM) to compose an answer in natural language, rather than returning a ranked list of links. Examples include:

• ChatGPT with web browsing enabled (OpenAI)
• Perplexity AI — retrieval-first, cites sources inline
• Google AI Overviews (formerly Search Generative Experience)
• Microsoft Copilot — powered by Bing and GPT-4
• Claude (Anthropic, with web tools)

These tools share a common architecture: they retrieve relevant content from the web using a mechanism known as Retrieval-Augmented Generation (RAG), introduced by Lewis et al. at Facebook AI Research in 2020 (arXiv:2005.11401). The AI fetches passages from external websites, combines them with its own knowledge, and generates a response. The websites it fetches from get cited — or not.

GEO: Optimising for That Citation

Generative Engine Optimisation (GEO) is the discipline of structuring your content so that AI engines select it as a cited source. The term was formalised by Aggarwal, Murahari, Rajpurohit et al. at Princeton University and IIT Delhi, presented at ACM KDD 2024 in Barcelona (arXiv:2311.09735).

Their key finding: content that includes statistics, citations from credible sources, and direct quotations can boost visibility in AI-generated responses by up to 40% compared to unoptimised content. That number comes from systematic evaluation across 10,000 queries on GEO-bench, validated on Perplexity.ai.

You may also encounter the term Answer Engine Optimisation (AEO) — both describe the same goal from slightly different angles. AEO emphasises optimising for voice assistants and direct-answer boxes; GEO emphasises the broader shift to generative AI as the primary search interface. Both are part of the same visibility strategy.

GEO vs Classic SEO: What Actually Changes

Classic SEO and GEO are not mutually exclusive — but they reward different content choices.

Factor	Classic SEO	Generative Engine Optimisation (GEO)
Primary goal	Rank high in a list of links	Be cited inside an AI-generated answer
Success metric	Position 1-10 in SERPs	Named as a source in AI response
Key ranking signals	Backlinks, domain authority, keywords	Factual clarity, citations, schema, E-E-A-T
Content format rewarded	Long-form keyword-rich pages	Direct definitions, FAQ blocks, structured data
Technical requirement	Fast site, mobile-friendly, sitemaps	JSON-LD schema (LocalBusiness, FAQPage, Review)
Measurement tool	Google Search Console, Ahrefs	Perplexity.ai testing, BrightEdge, Semrush AI
Time to effect	Weeks to months	Weeks (schema); months (authority)
Zero-click risk	High (featured snippets)	Very high — AI answers without sending traffic

Both disciplines share a foundation: a fast, technically sound, trustworthy website. If you are already investing in good SEO, GEO is the next layer — not a replacement. Explore the full picture in our visibility pillar.

How AI Engines Decide What to Cite

Understanding the selection process helps you act on it. Most generative engines follow a similar chain:

1. Query parsing. The AI interprets the user's intent — "best emergency plumber Leeds" is understood as a local service request, not an informational query.
2. Retrieval. The system fetches candidate documents (your website, review sites, trade directories) using a dense vector search.
3. Ranking for relevance and trust. Passages are scored. Content that states facts clearly, links to sources, and carries schema markup tends to score higher.
4. Generation. The LLM synthesises an answer and decides which sources to name. Sources are cited if they directly contributed to a claim in the response.

Your website competes at steps 2 and 3. The levers you control are content quality, structured data, and authority signals.

The Three Pillars of GEO for Trades

1. Structured Data and Schema Markup

Schema.org markup is the single highest-leverage GEO action for a trades business. It converts human-readable content into machine-readable facts that AI engines can pull into a response with confidence.

Priority schema types:

• LocalBusiness — or specific subtypes: PlumbingService, RoofingContractor, ElectricalContractor, HVACBusiness. Declares your business name, address, telephone, service area, and opening hours in a format AI engines treat as verified fact.
• FAQPage — each question-answer pair becomes a discrete, citable unit. Generative engines frequently pull FAQ content verbatim into responses.
• Review / AggregateRating — star ratings and review counts appear in AI-generated trust summaries.
• HowTo — step-by-step guides (for example, "how to know when to call a plumber") are routinely cited when users ask procedural questions.

Implement all schema as JSON-LD in the <head> of each relevant page. Validate with Google's Rich Results Test before publishing. The glossary contains plain-English definitions for every schema term if you are new to this.

2. Citable, Factual Content

The research from Aggarwal et al. is direct: content that cites sources, includes statistics, and provides quotations from credible sources outperforms vague, unsubstantiated copy in AI retrieval. For a trades business this means:

• Replace "we are experienced professionals" with "our Gas Safe registered engineers have completed over 800 boiler installations in West Yorkshire since 2018."
• Add a definitions section or glossary entry for every technical term — AI engines favour content that explains rather than assumes.
• Include data points with sources: "According to OFTEC, around 1.5 million UK homes use oil heating" is citable; "many UK homes use oil" is not.
• Structure answers to common questions as discrete FAQ blocks — one question, one direct answer, then elaboration. See our blog for worked examples across different trades.

3. E-E-A-T: Experience, Expertise, Authoritativeness, Trustworthiness

Google introduced E-E-A-T as a quality framework for human raters — and its logic has been absorbed into AI retrieval too. Content that signals genuine Experience and Expertise is preferred over generic copy.

For trades businesses, E-E-A-T signals include:

• Credentials stated on the page. Gas Safe registration number, NICEIC approval, CHAS accreditation, Checkatrade membership — list them with registration numbers, not just logos.
• Author attribution. Articles written by a named, qualified tradesperson carry more authority than anonymous content.
• Local specificity. Mentioning specific streets, boroughs, or postcodes you serve, alongside local regulations (Part P electrical work, Building Regulations Part L), signals genuine local experience.
• Third-party citations. Links to and from Which? Trusted Traders, Trustpilot, Google reviews, and trade bodies (Gas Safe Register, NICEIC, NHBC) strengthen trustworthiness signals.

Applied GEO: What to Do This Month

Practical steps, in order of impact:

1. Add LocalBusiness JSON-LD to your homepage and every service page. Include areaServed, hasOfferCatalog, telephone, and openingHoursSpecification. If you are a plumber, use PlumbingService as the @type.

2. Create a five-to-eight-question FAQ section on every service page. Answer each question in two to four sentences. Mark up with FAQPage schema. Browse the comparisons section to see how we structure question-led content across different topics.

3. Publish at least one authoritative guide per core service. A 1,500-word guide on "how to spot a boiler that needs replacing" written by a named Gas Safe engineer, citing HHIC or Energy Saving Trust statistics, will outperform five thin pages in AI retrieval.

4. Claim and complete every directory listing. Checkatrade, Rated People, TrustATrader, and Yell all feed data to AI engines. Consistent NAP (Name, Address, Phone) across all listings is a trust signal.

5. Respond to every Google review. Review responses are indexed and retrievable. A thoughtful response referencing the job type and location adds citable, locally relevant content.

6. Link internally with descriptive anchor text. Use "our emergency plumbing service in Manchester" rather than "click here." AI engines follow link context when assessing topical relevance. Our guide to how to show up in ChatGPT and Perplexity goes deeper on internal architecture.

For trades businesses specifically, the guide to marketing for plumbers in the UK covers how these principles apply to the full marketing stack, not just search.

What GEO Does Not Change

A few things worth being clear about:

• GEO does not replace local SEO. Google Maps, Google Business Profile, and local pack rankings remain critical — most AI engines surface Google Maps data for near-me queries. These are complementary channels, not alternatives.
• GEO is not about tricking AI. Stuffing pages with keywords, generating thin AI-written content, or fabricating credentials will not work. AI retrieval systems are trained to prefer depth and specificity over volume.
• Zero-click is a real risk. AI answers often satisfy the query without the user visiting your site. The counter-strategy is to be the business named in the answer, which drives direct searches and calls rather than clicks. Brand recognition at the AI layer is the new top of funnel.

The Takeaway

The shift to generative AI as a primary search interface is not coming — it is here. For trades businesses in the UK, the window to establish authority in AI-generated answers is open now, before the market saturates. The required actions are concrete and achievable: structured data, factual and cited content, verifiable credentials, and consistent local signals.

Start with schema markup on your homepage and one service page this week. Add a five-question FAQ block. Then build from there. The businesses that act early will be the ones AI engines cite — and the ones customers call.

For a broader framework, see our visibility pillar and the glossary for plain-English definitions of every term used in this article.