AI SEO for NZ businesses: get cited by ChatGPT, Perplexity and Google AI
AI Overviews have cut organic click-through on informational queries by 65%. Conversational AI search is now naming a short list of businesses in its answers. We restructure your content, schema and agent readiness so your business is the one being named, on Google AI Overviews, ChatGPT, Perplexity, Claude and Gemini. We are doing this on our own site, in public, as the proof.
Trusted by NZ businesses building for AI search since 2018








































What changed for NZ businesses already investing in SEO
If your organic traffic has dropped over the last 12 months and the rankings look mostly fine, the three things below are almost certainly the reason. None of them are about how the site looks.
Your organic clicks are dropping even as rankings hold
The 2026 data has it black and white. Organic click-through rate on queries that show an AI Overview has fallen from 1.76% to 0.61%, a 65% drop. Impressions stay flat, sometimes climb, and clicks quietly bleed away. Most NZ businesses watching their blog traffic crater have no idea the AI Overview is the reason.
Your brand never gets cited inside the AI answer
Someone asks ChatGPT "best web design agency Auckland" or Perplexity "how much does SEO cost in NZ" and the model lists three or four businesses by name. You are not one of them. Your content is too generic for extraction, your schema is thin, and the model has no reason to trust you over a US blog covering the same topic.
Your site is not built for agentic browsing
Google Lighthouse 13.3 now ships an Agentic Browsing audit. It checks WebMCP tool registration, accessibility-tree health, layout shift, and llms.txt presence. Most NZ business sites fail three of the four checks. The next wave of AI agents will book, compare, and buy on behalf of users, and the sites they can read will be the sites that get the booking.
Traditional SEO vs AI SEO
Traditional SEO still matters. AI tools with live search pull from the top-ranking pages, so you cannot skip the foundations. What is different is what you optimise on top of that. The six rows below are where the two disciplines actually diverge.
The pattern we see: NZ businesses who add the AI SEO layer in 2026 are the ones holding organic share as AI Overviews and AI Mode keep rolling out. Without it, the bleed compounds quarter over quarter.
Not sure where AI search is eating your leads?
Book a free 30-minute AI SEO audit. We run the Lighthouse Agentic Browsing audit, check how often your top queries trigger an AI Overview, check whether ChatGPT and Perplexity mention you by name on your topic, and pull your AI referral traffic from GA4. You walk away with a clear, ranked fix list.
Get my free AI SEO auditWhat AI SEO actually looks like as a service
Six work streams. Some of these are technical, some are content, some are off-site. The retainer mix depends on where the gaps are when we run the baseline.
AI Overviews extraction optimisation
Rewriting the opening 100-150 words of high-value pages so the answer is extractable. Adding FAQPage and Article schema where it is missing. Restructuring headings to match the questions buyers actually ask. The same playbook we wrote about in our 61% CTR drop post, applied to your top 20 pages.
LLM citation strategy
Definitional content, original data, named author bylines, and the kind of plain-English clarity that LLMs prefer to extract. Pulling your business into "best of" roundups, NZ directories, and press placements that show up in the query fan-out searches ChatGPT and Perplexity run live.
Schema and structured data expansion
A site-wide audit and expansion of structured data. Article on every blog post, FAQPage on every service and resource page, Person schema for the author, Service for what you sell, BreadcrumbList and sameAs links. Schema is now table stakes for AI extraction, not a nice-to-have.
WebMCP and agentic browsing readiness
Running the Lighthouse 13.3 Agentic Browsing audit on your site and fixing what it surfaces. Accessibility tree gaps, layout shift on click targets, llms.txt at the root, and WebMCP scoping on the transactional flows that matter. Future-important, started now so you are not retrofitting in 18 months.
AI visibility tracking and reporting
Monthly tracking of which queries trigger AI Overviews on your topics, which of those cite your content, and how much referral traffic you are pulling from chatgpt.com, perplexity.ai, copilot, gemini and others. A GA4 regex you can run yourself, plus our own tooling layered on top.
Content rewriting for AI extraction
A structured rewrite of your most valuable existing content. Direct answer first, question-style headings, short paragraphs, real numbers and sources, named author. Not new content for the sake of it. The same content, restructured so an LLM can pull a clean answer without hallucinating around your message.
Why NZ businesses choose Lucid for AI SEO
Most NZ agencies pitching AI SEO in 2026 are repackaging a 2022 SEO playbook with new vocabulary. The brand voice changes, the work does not. We have been actively building for AI search on our own site since early 2026, and we have published the data and the playbooks in the open as we have gone.
The blog posts on AI Overviews, the Lighthouse Agentic Browsing audit, AI SEO citation strategy, and content optimisation for AI search are all linked from this page. They are not lead-magnet fluff. Each one is structured the way we structure client content: direct answer in the first 100 words, named author, real data, FAQ schema, clean LLM extraction.
Our /glossary/ and /compare/ sections were built specifically as AI-extractable definitional content. We track our own AI referral traffic and citation share monthly. When we audit your site, we are running the same checks against your content that we run against ours. The work, the proof and the pricing are visible.
What this looks like for your business
Citation inside AI Overviews on the queries that matter
Schema, opening paragraphs and FAQ sections restructured so Google AI extracts answers from your pages, not your competitor's.
Brand mentions inside ChatGPT, Perplexity and Gemini answers
Citation footprint built across roundups, directories and consistent NAP data so LLMs treat you as a known, trusted entity in your category.
A clean Lighthouse Agentic Browsing pass
Accessibility tree gaps closed, layout shift cleaned up, llms.txt at root, and WebMCP considered on the transactional flows that matter. Three audits passed at once: AI readiness, accessibility and SEO.
AI referral traffic you can actually see
GA4 set up to surface chatgpt.com, perplexity.ai, copilot, gemini and others as named sources. Monthly reporting on citation share and AI traffic, not just keyword rankings.
Fast sites get cited. Slow sites get skipped.
AI Mode, Perplexity and ChatGPT search all run live fetches against your pages. They give a budget per page, and pages that take too long to render or fail Cumulative Layout Shift checks get quietly dropped from the result set. The Lighthouse Agentic Browsing audit calls this out explicitly: CLS has moved from "Core Web Vitals ranking factor" to "blocking issue" for AI agents.
The fix list is the same one technical SEO has been pushing for years, just with a sharper edge. Set width and height on every image, reserve space for embeds, avoid injecting content above existing content after first paint, ship clean accessible HTML. The new wave of AI agents reads your accessibility tree the same way screen readers do, so every accessibility win is also an AI win.
You can plug your current site into our free website speed impact calculator and see what slow load times and layout shift are costing you on both Google and the AI tools that depend on Google's index.
How we work with AI SEO clients
Four phases. You see the baseline, the priority fix list, the work as it happens, and the monthly reporting. No vague "AI strategy" deliverables.
AI visibility baseline
We run your site through the Lighthouse Agentic Browsing audit, pull a 30-day sample of your top queries to see which trigger an AI Overview, check ChatGPT, Perplexity and Google AI for citations on your topic, and review your GA4 for AI referral traffic. You walk away with a clear picture of where you actually stand.
Priority fix list
A short, ranked list of what to fix first. Usually a mix of schema gaps, accessibility tree issues, the opening 100 words of your top 10 pages, and llms.txt as a two-minute audit pass. We are not going to bill you for a five-figure llms.txt project. Mueller called it the new keywords meta tag for a reason.
Implementation and content rewrites
We do the schema work, the accessibility tree fixes, the content restructuring, the FAQ additions, the citation push to roundups and directories. You see drafts before anything ships. Where the work touches code we do it. Where it touches content we draft, you approve.
Monthly tracking and iteration
Monthly reporting on AI Overview citation share, AI referral traffic, and any traditional ranking changes. We keep rewriting opening paragraphs and expanding schema across the next tier of pages. AI search is moving every month. The retainer keeps you moving with it.
What clients say about working with us
Real NZ businesses we have worked with on web, SEO and content systems since 2018.
What is slow loading and layout shift costing you in AI citations?
AI agents drop slow and unstable pages from their result set. Plug your URL into our free website speed impact calculator to see the revenue and citation impact of every slow second.
Try the free speed impact calculatorRelated reading
The four AI SEO posts below are the playbook we wrote and the playbook we run, on our own site and our clients'.
AI SEO FAQ
The questions NZ business owners and marketing leads ask on the first AI SEO call. Real answers, no hand-waving.
Ask a QuestionReady to be the business AI search names?
Tell me about your site, your top organic pages, and where the traffic is leaking. I will run the Lighthouse Agentic baseline, check your citation share across ChatGPT, Perplexity and Google AI, and come back with a ranked, itemised fix list. No vague AI buzzwords, no five-figure llms.txt projects.