How to Get Found, Cited, and Chosen in the Age of AI Search

On 19 May 2026, Google’s VP of Search, Liz Reid, said four words that every marketer should have tattooed on their strategy deck: “Google Search is AI Search.”

It wasn’t a product announcement. It was a declaration. The company that built the modern internet, the one that gave rise to SEO, content marketing, and the entire ecosystem of click-based marketing, has officially renamed itself. It didn’t say AI is coming to search. It said search is AI.

Not a warning of what’s coming, but confirmation of what has already arrived. The question is no longer whether to adapt. It’s whether you’ve already started.

The numbers make the case on their own. Google AI Mode reached 1 billion monthly users in its first year, with queries more than doubling every quarter. The average AI Mode query now runs three times longer than a conventional search; people aren’t typing keywords anymore, they’re having conversations. And Google’s total search query volume just hit an all-time high, even as click-throughs to websites decline.

That’s the paradox defining this moment: more people are searching than ever before, and fewer of them are arriving at your website. The search engine is growing. Your traffic isn’t necessarily benefiting.

Generative Engine Optimisation (GEO)

This is the context Generative Engine Optimisation (GEO) needs to be understood in. It isn’t a niche technical discipline for SEO specialists, it’s the new front door to brand discovery, and most brands don’t have a key for it yet.

How AI Search Actually Works

Traditional search is a blue-link model. You type a query, Google matches it against an index of crawled pages, ranks them by authority signals, and hands you ten ranked links. You choose which one to click, visit the site, and form your own opinion.

Generative AI removes that step. By using large language models and natural language processing, the models deliver direct, summarised answers without the user ever really needing to click through, in the traditional sense.

When you ask a question in Google AI Mode, ChatGPT, or Perplexity, the AI model breaks your question into several sub-queries and searches each independently. A question like “best laptops for graphic designers under R20,000” becomes five or six parallel searches. It retrieves content from across the web, reads and assesses it for credibility and relevance, and synthesises a single direct answer. Then it picks two to five sources to cite visibly. Everything else it used stays invisible. The user gets a complete, conversational answer, and may never see your website URL.

There’s no position seven in this model. No page two. No second chance through a better meta description. The AI model either selects your brand as a cited source, or it doesn’t. The editorial decision has moved from the user to the AI.

The Grey Area

Between the traditional SEO world and the emerging GEO world, sits a zone of real strategic confusion, and most marketing teams are caught in it. Google Search Console shows rising impressions, so your content is being seen more than ever, while organic traffic stays flat or falls.

Firstly, what’s actually happening is that your content is feeding AI-generated answers the user never clicks through from. Your brand is influencing decisions you can’t see, doing work you can’t measure. Marketing teams reporting exclusively on sessions, pageviews, CTR, and bounce rate are flying blind in what’s bcoming the most important channel of the decade.

The second strategic error is treating SEO and GEO as competing priorities. They are not. They are deeply symbiotic. The same content and authority signals that power strong traditional SEO are the foundation GEO performance is built on. What’s changed is the ceiling: traditional SEO gets your content indexed. GEO-ready content gets your brand cited. That’s the difference between existing in the search ecosystem and winning commercial outcomes from it.

The third is a content volume trap. Generative AI has made it trivially easy to produce content at scale, and many brands have responded by flooding the zone. But AI search systems are becoming more selective. They reward specificity, expertise, and verifiable authority, not volume. A brand publishing four deeply researched, credibly sourced pieces a month will consistently outperform those not investing accordingly.

The fourth is single-channel vulnerability. Most brands built their visibility strategy around Google organic and Google Ads. That duopoly is fracturing: ChatGPT now accounts for roughly 77% of AI-driven referral traffic globally, Perplexity around 15%, and Copilot is embedded in hundreds of millions of Microsoft workflows. Each platform has a different retrieval architecture and different citation logic. Optimising for Google alone leaves most of the AI discovery landscape unaddressed.

The South African Angle

It’s tempting for South African marketers to treat GEO as a developed-market problem that arrives here later. The data says otherwise. 81% of South African consumers already use AI tools like ChatGPT, above the global average of 68%.

The professionally active, higher-income, mobile-first segment that represents most B2B and premium B2C buyers is already using AI as its default research interface.

If your brand isn’t in those AI-generated answers, you’re not in those consideration sets. Most South African brands haven’t started building for GEO yet, which means the competitive window is still wide open.

The brands that start now will be the ones AI is citing by 2027 and 2028, when the market has fully consolidated around AI-first search behaviour. Brands that build it now will have a structural advantage latecomers may not be able to buy their way into.

Measuring What Actually Matters Now

If your reporting stack only shows traditional metrics, you’ll look like you’re losing ground even when you’re winning.

It’s time to retire organic traffic volume, keyword rankings, and click-through rate as primary KPIs on their own, and add AI citation frequency, share of answer, AI referral traffic quality, and brand representation accuracy alongside them.

Some of this can be done manually, regularly searching your priority queries in ChatGPT, Perplexity, and Google AI Mode and tracking whether you’re cited, or watching the Search Console impressions-vs-clicks divergence as a signal your content is feeding AI answers. But manual audits don’t scale well against a landscape with this many platforms and this much query volume.

A growing number of dedicated tools, from global players, to local platforms such as Signal Index https://www.signalindex.co.za/ — now offer a way to measure overall how visible your brand is inside AI-generated answers and discovery systems, rather than tracking one query at a time.