Practical Tips to Dominate GEO, AIO and SEO in New Zealand in 2025 and Beyond by The Optimisers

With AI-driven search reshaping digital visibility, Kiwi marketers must evolve fast. Richard Conway, Managing Director of The Optimisers and Pure SEO, shares key insights on how New Zealand businesses can stay competitive in 2025 and beyond. Read below to know more.

Since speaking at the Digital Day Out recently, I have had lots of CMOs, Marketing Managers and business owners reach out to me to understand how to appear in the different LLMs (ChatGPT, Perplexity, Claude etc). Because of the sheer volume of questions, I have decided to document some of the more easily actionable activities for the Marketing Association’s resource hub.

Traditional SEO has been and is still central to digital marketing strategies in New Zealand, the rise of Generative Engine Optimisation (GEO), or interchangeably Answer Engine Optimisation (AEO), creates a new dynamic and opportunity for Kiwi companies.

On top of this, our own data at The Optimisers - www.theoptimisers.com indicates that almost 30% of queries that would traditionally bring traffic to websites in the top and middle of the funnel are now being serviced within Google (AI Overviews) or the LLMs themselves, giving rise to a reduction in traffic and a huge increase in zero click searches.

This reduction in website traffic and the way users now consume the information has also made attribution much harder. It almost feels like we are going back a decade in respect of the ability to truly understand the user journey and how they interact with brands.

AI Overviews

Google has integrated AI within search in New Zealand through AI Overviews. In the US and India, their latest iteration of this is AI Mode, this will be rolled out globally later this year or early next year. With my own experiments, the outcomes are much better than AI Overviews – multi-modal (text, image, audio, video) and more accurate.

The first thing to consider is why AI overviews selects certain sources and favours certain websites. The general consensus across my mastermind group and the data that has been published is that brands with top web mentions earn up to 10X more AI mentions, this was confirmed by Brightedge Labs in May):

Having delved into this in a far deeper detail at The Optimisers we ascertain (predominantly with NZ and AU data) it has also become apparent that there is a bias towards mentions in top tier media (not media outlets known for accepting a high volume of guest contributions) and has favourable discussions on social sites such as Reddit and other social platforms. Interestingly enough there seems to be a negative correlation between negative comments and appearance in AI Overviews – your reputation is important!

Resource: https://ahrefs.com/blog/ai-overview-brand-correlation

Technical SEO / AIO / GEO are still critical, things like structured data (schema markup), page load speeds and quality of information (EEAT -https://www.searchenginejournal.com/google-e-e-a-t-how-to-demonstrate-first-hand-experience/474446/) are still absolutely non-negotiable in respect of visibility.

The biggest shift in search can be seen by looking at the patents related to AI search. In this study of the six most critical patents, Michael King highlights how the future of search is likely to be deterministic rather than probabilistic (https://searchengineland.com/how-ai-mode-ai-overviews-work-patents-456346)

(A deterministic model is like a recipe: if you follow the exact same steps with the same ingredients, you always get the same result — no surprises.

A probabilistic model, on the other hand, is like rolling dice: even if you do the same thing every time, the outcome can change because there’s some built-in randomness.

So:

Deterministic = predictable and repeatable

Probabilistic = allows for chance and uncertainty.)

I.e. We would create content, Google would crawl and parse it, then surface specific elements in the SERPs based on you gave it. But it wasn’t understanding the information—it was simply extracting.

We still rely on strong fundamentals—solid content, clean architecture, accessible links—but now, our visibility also depends on a complex, often invisible reasoning process. LLMs don’t just extract; they interpret, weigh relevance dynamically, and factor in past user interactions. They also require the ability to deal with more complex, conversation queries rather than individual search terms or phrases.

Other LLMs – Chat GPT, Perplexity, Claude

As we know, these new LLMs don’t have the data or search legacy that Google has. This can be interpreted as either an advantage or a disadvantage (it is both in my opinion).

A recent study by Semrush indicated a significant overlap between Google’s top 10 organic rankings and LLM citations.

With all the platforms there has been a strong correlation between having a strong SEO presence (top 10 rankings) and AI search citations, I would expect this to continue (hence traditional SEO is still critical.

Summary

Whilst the landscape is still changing all the time and our team at The Optimisers are monitoring, adapting and optimising clients' websites, I have put together the following table to assist Kiwi businesses navigate the current situation when considering how they format and write their content.

Resource: https://pureseo.com/nz/blog/geo-vs-seo-ai-search-engines

Over the past 18 months as an agency, we have invested millions into research around LLMs and search on behalf of large New Zealand businesses, SaaS and public companies to ensure that they adapt their strategies to benefit from this new wave of search. Whilst attribution is an increasing problem our specialists work on ensuring our family of clients have access to the best data and latest recommendations. My mantra is ‘always lead, never follow!’

Source: Richard Conway, 30 July 2025