When we ran thirteen live Australian trades searches in July 2026, not one returned an AI Overview. Every single one returned a local map pack. If an agency has told you that AI Overviews are about to take your building enquiries, our own data says otherwise.
That does not mean AI search is irrelevant to builders. It means the risk is somewhere other than where most of the industry is pointing. This guide covers what our research actually shows, where the real AI opportunity sits for a building company, and what to do about it.
Key takeaways
- Zero of thirteen Australian trades searches returned an AI Overview. All thirteen returned a map pack.
- The real AI question for builders is being recommended in ChatGPT, not appearing in an AI Overview.
- When AI does recommend, it cites third parties. Five of six sources in the one AI Overview we could inspect were not businesses.
- Reddit appeared in the top 10 on 38.5% of trades searches, and those threads feed AI recommendations.
- The work that wins AI recommendations is the same work that wins the map pack, which makes this cheap to justify.
Do builders need to worry about AI Overviews?
Not currently, on the searches that generate work. In our study of 65 live Australian service searches conducted on 31 July 2026, AI Overviews appeared on 4.6% overall and on zero of the thirteen trades searches. A local map pack appeared on 100% of trades searches and People Also Ask on 92.3%. Google serves an answer box when a query needs explaining and a map pack when someone wants to hire. Building searches are hiring searches.
There is a real caveat. Our study covered commercial local searches like “builder melbourne”, not informational ones like “how much does it cost to build a house”. AI Overviews are considerably more common on the second kind, and builders do get informational traffic. So the honest position is that AI Overviews threaten your blog traffic more than your enquiry traffic.
Full method and limitations are in our AI Overviews study.
So what is the real AI question for builders?
Whether ChatGPT names your company when someone asks it to recommend a builder.
That is a different question from AI Overviews and it has a different answer. Someone planning a $600,000 build now routinely asks an AI assistant who they should talk to, how to tell a good builder from a bad one, and what questions to ask. Those conversations happen months before any form gets filled in, and you are either in them or you are not.
How do AI assistants decide which builder to recommend?
By reading public sources and reconciling them. Your website establishes what you build and where. Your Google Business Profile supplies structured facts. Reviews, association registers, directories, local media and community discussion corroborate it. Where those agree, the system has enough confidence to name you.
The evidence for how heavily third parties matter is in our own study. In the one AI Overview that returned its sources, six were cited and only one was a business selling the service. The other five were government, community, comparison and tool sites.
For a builder that means being recommended requires being written about, not writing about yourself.
What should a builder actually do?
Five things, in order of return.
1. Finish your Google Business Profile
The densest source of structured facts about your business that exists, and it drives the map pack that appeared on every trades search we tested. Categories set precisely, every service with its own description, real service areas, project photos, and the questions and answers section populated.
This single job serves the map pack and AI recommendations at once, which is why it sits first. Our SEO guide for builders covers the detail.
2. Get reviews that name the project and the suburb
“Great builder” is not retrievable by anything. “Built our double-storey extension in Coburg, on budget and two weeks early” maps directly to the queries you want to win and gives an AI system a specific fact to repeat.
3. Fix your entity data everywhere
Your trading name, address, phone, licence number and service areas identical across your website, Google Business Profile, association listings, industry directories and any builder registers you appear on. Every contradiction reduces confidence, and confidence is what decides whether you get named.
Builders are unusually prone to this because licence details, ABNs and trading names change as companies restructure while old listings persist.
4. Answer the questions buyers ask before they choose
The genuinely useful AI-facing content for a builder is not more pages about your quality workmanship. It is direct answers to the questions people ask an assistant during research:
- How much does it cost to build a house in [city]?
- How do I choose a builder?
- What should be in a fixed price building contract?
- What is the difference between a custom builder and a project builder?
- How long does a knockdown rebuild take?
- What questions should I ask a builder before signing?
Each needs a question-shaped heading with a direct 40 to 100 word answer underneath, written so it survives being quoted alone. That last question is the one worth owning outright, because it is exactly what people ask an AI before their first meeting with you.
5. Earn independent mentions
Master Builders or HIA member listings completed properly, local media, industry awards, genuine community presence. Reddit appeared in the top 10 on 38.5% of the trades searches we tested and those threads are read by AI assistants when someone asks for a recommendation.
You cannot manufacture this. Paying for reviews or commissioning forum posts breaches Australian Consumer Law and the exposure sits with your business. What you can do is be worth recommending and be present where builders are discussed. Our guide to the sources AI assistants cite covers the mechanics.
Free Download
The Builder’s Growth Playbook
- The system builders use to turn $50k in ad spend into $5M to $15M in contracted work
- Why the best clients in your market are choosing builders worse than you
- The five mistakes that destroy the result before it has a chance to work
How do you test where you stand?
Fifteen minutes, free.
- Open ChatGPT and Perplexity in fresh, logged-out sessions.
- Ask what a real customer would ask: “Who is the best custom home builder in [your city]?”, “I want to build an extension in [suburb], who should I talk to?”, “[Your company] reviews”.
- Record whether you were named, who was named instead, and which sources were cited.
- Perplexity shows its citations, so you can trace exactly which source is feeding a competitor’s recommendation.
- Repeat monthly.
Most builders discover the problem is absence rather than being beaten. The system has nothing specific to say about them and names whoever it can describe.
What we recommend
Do not buy AEO for builders as a defensive purchase against AI Overviews. Our own data says that threat is not currently present on the searches that produce building enquiries, and any agency selling it on that basis should be asked for evidence.
Buy it, or do it yourself, because the work that wins AI recommendations is the same work that wins the map pack: a complete profile, specific reviews, consistent entity data and content that answers real questions. You are getting two outcomes from one job, and the map pack alone justifies it.
Run the fifteen-minute test first. If AI already names you, spend the effort elsewhere.
Frequently asked questions
Do AI Overviews affect builders in Australia?
Not on the searches that generate work. Across thirteen Australian trades searches in our July 2026 study, AI Overviews appeared zero times while a local map pack appeared on all thirteen. They are more common on informational questions like construction cost queries, so they affect blog traffic more than enquiry traffic.
How do I get my building company recommended by ChatGPT?
Complete your Google Business Profile, generate reviews that name the project type and suburb, keep your business details identical across every directory and register, and answer the questions buyers ask during research. AI assistants reconcile public sources, so the work is making those sources accurate and specific.
Is AEO worth paying for as a builder?
It is worth doing, and the justification should not rest on AI Overviews. The same work drives your map pack visibility, which appeared above the first organic result on every trades search we tested. If a provider sells AEO to you purely as AI Overview protection, ask them for the trades data.
What questions do people ask AI before hiring a builder?
Commonly what it costs to build in their city, how to choose between builders, what a fixed price contract should contain, how long a build takes, and what to ask before signing. Answering those directly on your site is the highest-value AI-facing content a builder can publish.
Does Reddit matter for building companies?
It appeared in the top 10 organic results on 38.5% of the Australian trades searches we tested, and those threads are among the sources AI assistants read. You cannot safely manufacture presence there. Being genuinely worth recommending and visible in your local community is the only durable route.
How long before AI starts recommending my building company?
Google Business Profile and directory corrections can influence AI answers within weeks, because most assistants run live web searches. Review volume and independent mentions build over one to three months. Test monthly and judge on a quarter.
Want to know what AI says about your building company?
We run AI visibility audits for Australian builders: prompt testing across ChatGPT, Perplexity and Gemini, source tracing for every competitor recommendation, and a prioritised fix list.
See our construction marketing services and AEO services, or book a free audit.
Sources
- Uprise Digital (2026), AI Overviews in Australian Service Search
- Google Search Central, AI features and your website
- Australian Competition and Consumer Commission, guidance on reviews and testimonials
