Google AI Mode SEO for Trades: How to Get Cited, Compared and Recommended
The opportunity is bigger than ranking for “roofer near me”. AI Mode can interpret the job, compare evidence and recommend businesses that appear to fit the brief.
Imagine a homeowner asks Google: “I need a roofer near Guildford for a leaking valley on a 1930s house. I want someone experienced with leadwork, with evidence of similar jobs and good local reviews.”
Traditional SEO tends to reduce that to a keyword such as roofer Guildford. Google AI Mode does not have to. It can break the request into related searches, retrieve evidence from different pages and sources, and assemble an answer around the user’s actual problem.
Google calls this query fan-out: AI Mode can issue multiple related searches across subtopics and sources before producing its response. For trades, that changes the optimisation target. The winning page is not necessarily the one that repeats the main keyword most effectively. It is the business that gives Google the clearest evidence that it is relevant to the specific job.
Google says AI Mode has surpassed one billion monthly users globally. In the UK, it has been available since July 2025, so this is no longer a future-search thought experiment for British trade businesses.
Being Cited Is Not the Same as Being Recommended
This is one of the most important distinctions in AI search. In a 2026 analysis of B2B software queries, Lily Ray found that Google AI Overviews often cited brands’ own “best” listicles while leaving those brands out of the recommendation set. In 69% of those citation cases, the cited brand was not recommended.
That study was not about local trades, so it should not be treated as a local ranking factor. But it exposes a useful principle: a page can be useful evidence without the business behind it being judged the best fit for the user.
Can Google use your fact?
Your page contains a clear passage, project detail or answer relevant to the question.
Can Google distinguish you?
Your experience, location, project proof and third-party signals make you meaningfully different from alternatives.
Do you fit the brief?
Your service, location, evidence and reputation appear to match the customer’s particular need.
AI Mode Turns One Trade Search Into Several Evidence Checks
Google says AI Mode is particularly useful for exploration, comparisons and nuanced questions. It may use query fan-out to find supporting pages that would never have ranked for the user’s exact wording.
Take the Guildford roofing example. Behind one question, useful evidence could come from retrieval steps around:
- roofers serving Guildford
- lead valley repair
- 1930s or period-property roofing experience
- examples of similar roof repairs
- local reviews and reputation
- guarantees, response times or service details
This is why AI SEO for trades should not mean creating hundreds of thin pages for every long-tail query. It should mean creating a website and wider online presence containing specific, corroborated facts about the work you actually do.
What Search Experts Are Saying
“Create clear, self-contained, atomic answers that stand on their own.”
“The query fan-out technique…is the big thing that I think we need to be thinking about in our space.”
“Think of local search as a digital ‘word-of-mouth’ system.”
The common thread is useful for trades: specificity beats generic SEO copy. A page saying “we provide high-quality roofing solutions at competitive prices” gives an AI system almost nothing distinctive to work with.
The Construct Virtual PROOF Stack for AI Search
For a trade business to move from merely appearing online to being a plausible AI recommendation, we use five evidence layers.
The PROOF Stack
What do you actually do, for whom, and in what situations?
Jobs completed, materials, problems solved, photos, outcomes and constraints.
Areas served, property types, local conditions and genuinely relevant service information.
Reviews, trade bodies, suppliers, directories, press, partners and consistent business data.
Accurate Business Profile details, indexable text, current services, images and internal links.
P — Precise services
Do not make Google infer what “all aspects of roofing” means. A useful roofer page might explicitly cover slate repair, tiled roofs, lead valleys, flat-roof repairs, chimney flashing and storm damage—assuming those are services the business genuinely provides.
R — Real project evidence
Project pages are one of the best AI-search assets a trade business can own because they create facts competitors cannot copy.
Evidence: “On this 1930s Guildford semi, water was entering where the rear roof met the lead valley. We lifted the surrounding tiles, replaced the failed leadwork and reinstated the original tiles where serviceable.”
The second version contains entities, location, building type, problem, method and material. It can answer several related questions without being written for a robot.
O — On-page local context
A London electrician, rural heating engineer and coastal roofer face different buildings, customers and recurring problems. Useful local SEO explains those differences rather than swapping a suburb name into the same template.
O — Outside corroboration
Your website is a claim you make about yourself. Reviews, relevant professional memberships, supplier listings, local press, project partners and other independent references can help corroborate the entity behind those claims.
F — Fresh, crawlable facts
Google’s guidance for AI features still comes back to core SEO: allow crawling, make important content available as text, use sensible internal linking, keep Business Profile information current and ensure structured data matches visible content.
What This Looks Like Across Different Trades
| Trade | Weak evidence | AI-search evidence worth creating |
|---|---|---|
| Roofer | “Roof repairs Surrey” | Case studies showing roof type, fault, material, area, repair method and before/after photography. |
| Plumber / heating | “Reliable local plumber” | Pages separating emergency leaks, heating faults, bathroom work and recurring property-specific problems, supported by real jobs. |
| Electrician | “Domestic and commercial electrical services” | Specific evidence around consumer-unit upgrades, rewires, fault finding, EV charging or commercial work actually completed. |
| Builder | “Extensions and renovations” | Projects explaining property type, access constraints, structural changes, programme, scope and finished result. |
| HVAC contractor | “Commercial air conditioning experts” | Project evidence by building type, system type, maintenance challenge, replacement scope and service coverage. |
| Specialist contractor | “Fire protection solutions” | Precise systems, substrates, project environments, accreditations where applicable, application evidence and completed-project examples. |
The Citation → Comparison → Recommendation Ladder
Can Google use your fact?
Clear service descriptions, concise answers, project facts and useful explanations can become supporting evidence.
Can Google distinguish you?
Specific expertise, locations, property types, project evidence and third-party proof help separate businesses that otherwise look identical.
Do you fit the brief?
Recommendation requires a combination of relevance, local fit, reputation and enough trustworthy evidence for the user’s particular request.
This is why a business can rank organically yet remain weak in an AI comparison. A conventional service page may be relevant to “roofer Guildford”, but provide too little evidence for “roofer experienced with lead valleys on older homes near Guildford”.
Seven AI SEO Changes I Would Make to a Trade Website Now
- Rewrite vague service copy into specific service evidence. Explain the jobs you do, recurring problems, materials or systems involved, who the service is for and where it is relevant.
- Publish fewer, better project case studies. Include the situation, location, problem, work completed, constraints, result and original photography.
- Add comparison information customers genuinely need. For example: repair versus replacement, different material options, what affects cost or which service fits which problem.
- Strengthen local entity consistency. Keep your Google Business Profile, website contact details, service areas and major third-party profiles accurate and aligned.
- Use original images as evidence. Google says images and video can create additional opportunities in generative AI search. Photograph actual work, not only finished glamour shots.
- Build contextual internal links. Connect service pages to relevant projects, technical articles, locations and proof.
- Look beyond your own site. Earn legitimate reviews, supplier or trade-body references, local mentions and project-partner citations rather than manufacturing dozens of low-value directory listings.
If you already have traditional trade SEO in place, this should feel evolutionary rather than revolutionary. Our guide to SEO vs GEO vs LLMO for construction companies explains why the fundamentals have not disappeared, while structuring a trades website for AI search goes deeper into site architecture.
What Not to Do
- Do not create 100 near-identical location pages and call it AI SEO.
- Do not add unsupported statistics because “LLMs like data”.
- Do not invent project experience, accreditations or service capabilities.
- Do not expect FAQ schema, an
llms.txtfile or a new schema type to make Google cite you. - Do not optimise sentences only for extraction until the page becomes unpleasant for customers to read.
How Do You Measure Whether This Is Working?
Google has started rolling out dedicated Generative AI performance reports in Search Console to a subset of websites. The reports can show which pages received impressions in generative AI features, together with country, device and date data.
If the report is available in your Search Console account, watch which service and project URLs appear, whether visibility grows after adding project evidence, and which page types earn AI impressions.
It is also worth manually testing high-value customer scenarios. Don’t only ask “best roofer Guildford”. Test the prompts a real customer might use when the job has constraints.
“Recommend roofing companies around Guildford that show evidence of repairing lead valleys on older tiled or slate homes. Explain why each business appears relevant and cite the supporting sources.”
Run variants. AI answers can change, so look for patterns rather than treating one response as a ranking report.
Construct Virtual’s AI search visibility audit guide covers the auditing process, while our AI Search Visibility service is designed for businesses that want this measured systematically.
The Bottom Line
Google AI Mode does not make SEO irrelevant. It makes thin SEO easier to expose.
If two electricians have similar titles, similar service pages and similar reviews, there is little for an AI system—or a customer—to distinguish. If one has precise service pages, detailed local projects, original photos, consistent business information and independent evidence, the comparison becomes much easier.
Build the evidence that proves you are relevant.
For a broader strategy, see how builders can rank in AI search. If you want this applied commercially, our AI SEO for Trades service combines traditional SEO, AI-search visibility and evidence-led content development.
Sources & Further Reading
- Google Search Central: AI features and your website
- Google Search Central: Optimising for generative AI features
- Google Search Central: Generative AI performance reports in Search Console
- iPullRank: AI Mode, query fan-out and user context
- Lily Ray / Amsive: GEO, AEO, LLMO — separating fact from fiction
- Search Engine Land: Lily Ray’s 2026 citation-vs-recommendation analysis
- Darren Shaw: Localised AI-search optimisation playbook
Can AI Search Understand Why Customers Should Choose You?
Construct Virtual helps UK trades turn real services, projects and local expertise into search evidence that works across traditional Google results and emerging AI search experiences.