When a customer asks ChatGPT or Google AI Mode which brand to buy from, the answer isn’t built from your website alone. AI engines lean on a surprisingly small pool of third-party sources (listicles, review sites, directories, and comparison articles) and they keep going back to the same ones.
If your brand isn’t on those pages, it isn’t in the answer. Unlike traditional search engines, which surface links from your own pages, AI-generated answers pull from the wider web, so brand visibility is now shaped by what others say about you.
That’s why a big share of our GEO (Generative Engine Optimisation) work for clients happens off-site.
Here’s how we track your AI search presence.
1. Start with the questions buyers actually ask
We build a set of 20–50 bottom-of-funnel prompts covering the relevant queries in your category: the questions people ask when they’re ready to choose, not just browse.
For an ecommerce client, that might be “Where’s the best place to buy outdoor furniture online in New Zealand?” rather than “What is teak?”
We track those prompts against your key competitors for competitive benchmarking, running the same prompt across different answer engines for at least a few weeks before drawing conclusions. AI responses vary, so you need a baseline to work with. By monitoring your brand mention frequency and share of voice, we can measure your brand’s presence in AI-powered search results.
2. Diagnose the problem by each AI engine
Where you’re weak tells us why. If you’re missing only from Google AI Overviews, it’s usually a traditional Google ranking problem, because those AI-generated recommendations draw heavily on what already ranks. For local businesses, your presence on Google Maps is vital; make sure your Google Business Profile is set up correctly and encourage your customers to leave reviews.
If ChatGPT and other AI assistants ignore you, your brand is probably missing from key directories, or has less presence on news sites and forums than competitors.
And if you’re absent on one topic across every AI tool, that’s a content gap. You might offer a certain service or product, but unless you dedicate a page with quality content to it, AI won’t pick that up.
Each AI brand issue points to a different fix, which stops budget going on the wrong one.
3. Map the sources that shape the AI search results
Next we identify which domains and pages AI engines cite most often in your category, and classify them: directories and review sites, editorial media, community forums, and competitor-owned content.
A proper competitor analysis makes this step much sharper. One pattern we watch for is a competitor whose own comparison pages or listicles are being treated as a category reference. That alone can explain most of a leader’s advantage.
4. Get onto the pages AI models already trust
From that map, we build an outreach list of pages where at least two competitors are named and you’re not, ranked by how often AI engines retrieve them. Competitor-owned pages are excluded (their “X vs Y” article will never add you).
Then we come back with a list of the highest-value directories and editorial media to reach out to, since those are the pages where potential customers and AI engines both look.
5. Trace negative claims back to their source
AI-generated responses repeat objections about brands: “expensive”, “slow shipping”, “limited range”. We trace each one to the page it came from and sort it, using light sentiment analysis to spot patterns.
If it’s outdated, we get the source corrected. If it’s a competitor’s framing, we create content that answers it. If it’s accurate, that’s useful product feedback. Before labelling anything outdated, we check real customer reviews. If customers are still saying it, the issue isn’t the source.
6. Clean up your listings
Directories, review platforms and social profiles all feed A-generated responses. Unclaimed, duplicated or inconsistent listings confuse the large language models and lead to inaccurate claims alongside brand mentions. We audit and fix name, description, category and pricing language, plus structured data, so every profile tells the same story for AI systems.
Most businesses have more control over how they show up in AI platforms than they realise. The key isn’t more information, it’s tidying up what AI already has. That might be an unclaimed listing, an old review quoting last year’s pricing, or a ‘best of’ list your competitors are on and you’re not. It’s all fixable. The daunting part is the scale, so start small, because it all adds up. The only thing worse than being invisible in AI results is AI finally showing you with the wrong information.
Why AI visibility matters for New Zealand businesses
In a market the size of New Zealand, the pool of local sources AI engines can draw on is relatively small.
That cuts both ways. A handful of NZ listicles, directories and review sites can carry real weight, so getting onto them can move the needle quickly. But if your competitors got there first, they’ll keep appearing in AI-generated answers until you focus on your own AI search visibility.
The 90-day plan for AI search visibility
We turn the audit into three workstreams: directory fixes, outreach, and correcting objections at their source, which also feeds your content strategy.
At day 90, we rerun the audit with exactly the same prompts, engines and competitors, so you can see what’s actually changed in AI visibility performance: visibility scores, share of voice and AI-driven traffic. We tag referrals in Google Analytics so the traffic side is easy to verify, and the audit becomes a measurement framework you can reuse.
Then we do it again. The sources AI systems rely on shift over time, so tracking brand mentions works best as an ongoing AI visibility strategy.
Find out your brand visibility in AI systems
Curious whether ChatGPT and Google AI Mode recommend you, or your competitors? We’re happy to run a free AI visibility review and show you where your brand appears and where it doesn’t. Get in touch at:
