The rise of GEO and the new rules of trust
By Jeremy Crowder, Managing Partner, Dialogue
The way people discover businesses is changing.
For years, the goal was relatively straightforward: rank well on Google, get people onto your website and make sure they could find the information they needed once they arrived.
AI is changing that journey.
At our recent The Space event, Jeremy Crowder, Managing Partner at Dialogue, explored what this shift means for businesses and brands, and why Generative Engine Optimisation (GEO) is becoming an increasingly important part of being found online.
His message? Search hasn’t disappeared. It’s becoming answers.
From storytelling to AI discovery
Jeremy’s career in communications started with a fascination for advertising and the ability to tell a complete story in 30 or 60 seconds. That eventually led him into strategy, brand management and communications, with an early lesson that has remained relevant throughout his career: understand what makes a business distinctive and communicate it clearly.
The principles of good communication haven’t necessarily changed. But the places where people discover, research and form opinions about brands have.
And AI is quickly becoming one of those places.
Search is becoming answers
Traditional search has generally been built around a simple behaviour: someone asks a question, receives a list of results and clicks through to find the answer. That last step is increasingly disappearing.
As Jeremy explained, people are still searching for information, but they can now get the answer they need without ever visiting the websites that provided it.
“Search did not get worse. It just stopped being search.
It became answers.”
That changes the challenge for businesses.
It’s no longer only about whether your website ranks for a particular search term. It’s also about whether your business, expertise and content form part of the answer an AI tool gives someone.
So, what exactly is GEO?
Generative Engine Optimisation, or GEO, is about improving the likelihood that your business or content is referenced when someone asks an AI platform a relevant question.
Jeremy described it simply as “being quoted when somebody else asks a question.”
SEO still matters. But instead of optimising solely to appear prominently in a list of search results, businesses now also need to think about how AI systems understand who they are, what they do and whether they are a credible source on a particular subject.
This introduces another important concept: entities.
An entity is essentially a person, company or organisation that an AI system can recognise as a distinct thing. If information about that entity is inconsistent, sparse or easily confused with something else, the system can struggle to understand it accurately.
And, as Jeremy pointed out, AI doesn’t necessarily know when it doesn’t know.
You can't write your own reference
One of the biggest themes of Jeremy’s talk was the difference between what a business says about itself and what other credible sources say about it.
Your website is important. It tells the world who you are, what you do and what expertise you have.
But third-party sources provide something your own website can’t: external validation.
Jeremy compared the difference to management accounts versus an independent audit. In his analogy, your website is what you say about yourself. Earned media, reviews, independent articles and other credible third-party mentions are the audit.
As he put it:
“You cannot write your own reference. AI trusts the audit, not the management accounts.”
Jeremy shared research from his presentation analysing millions of links cited across AI platforms, which indicated that the large majority of AI citations came from earned media rather than paid advertising or press releases distributed through newswires.
For businesses, that makes credibility outside your own channels increasingly important.
South African businesses have an opportunity
This shift could present a particularly interesting opportunity for South African businesses.
Jeremy’s argument is that many local categories simply don’t have the same depth of online content as their counterparts in larger markets.
AI models can only work with the information available to them. In categories where relatively little useful South African content exists, there may be fewer sources competing to become part of an AI-generated answer.
As Jeremy explained:
“In a category with almost nothing in South Africa, whoever publishes something genuinely useful becomes the source.”
Think about the kinds of questions potential customers might already be asking AI:
Who are the best Xero accountants in Cape Town?
What should I look for in an outsourced CFO in South Africa?
Do I need to register for VAT?
How do I choose a private bank?
These aren’t necessarily searches for a particular company. They’re searches for an answer.
The opportunity for businesses is to understand the questions their customers are asking and create genuinely useful content that helps answer them.
And Jeremy’s view is that, in South Africa, many of those spaces are still relatively open.
The new rules of trust
Being discoverable is only one part of the equation. The other is trust.
Jeremy explored how people increasingly rely on recommendations and information from sources they already trust, whether that’s friends and family, journalists, creators, industry experts or other third parties.
AI introduces another layer.
Information published somewhere online can be picked up by an AI system and repeated back to a user in the confident, neutral voice of an assistant.
Jeremy described this rather memorably as AI becoming a “trust laundering machine.”
Someone publishes something. The machine reads it. And that information can eventually be delivered to the user as an answer.
That makes the quality and accuracy of the information surrounding your brand increasingly important.
Credibility, in Jeremy’s words, is no longer simply something a business builds and holds. It can be linked to you by the sources your audience — and increasingly the machines they use — consider credible.
AI is replacing the first meeting
For accountants, advisors and other professional services businesses, one of Jeremy’s most interesting observations was about what happens before a potential client contacts you.
People are already using AI to ask questions, understand terminology, compare options and seek advice.
That means they may arrive at their first conversation with a professional having already formed an opinion about what they need.
Sometimes that information will be useful. Sometimes it may be incomplete or wrong.
As Jeremy put it:
“AI is not replacing the advisor. It is replacing the first meeting.”
The role of the expert hasn’t disappeared. But the starting point of the conversation is changing.
For businesses, the strategic response isn’t necessarily to resist that behaviour. It’s to think about how your expertise can become one of the sources informing it.
What can businesses do about it?
The encouraging part of Jeremy’s talk was that improving your visibility doesn’t necessarily require an enormous AI strategy or expensive technology.
There are some practical places to start.
See what AI already says about you. Search the questions a potential client might ask when looking for a business like yours. Jeremy recommends logging out of your AI accounts, starting a fresh chat and trying the same prompts across different tools. Don’t lead the answer by mentioning your company. See whether you’re named, how you’re described and which competitors appear instead.
Get your basic information right. Fix and update your Google Business Profile and make sure information about your business is consistent wherever it appears online.
Keep your people consistent too. Team biographies, job titles, photos and other information should align between your website, LinkedIn and other online sources. Jeremy emphasised that inconsistencies can make it harder for machines to connect information confidently.
Encourage public reviews. Happy clients talking about their experience creates third-party information about your business rather than relying entirely on what you say about yourself.
Answer the questions your clients actually ask. Look at the conversations that happen before someone becomes a client. What are they trying to understand? What do they search for? What questions come up repeatedly? Those questions can become useful website pages and articles.
Put names and faces to your expertise. Make it clear who the people behind your business are and what they know.
Build third-party credibility. Look for opportunities for your experts to contribute to relevant publications, podcasts and other credible platforms.
Create something original. If you have useful proprietary data, insights or research, publish it. Original information gives other people — and potentially AI systems — something valuable to reference.
Build a consistent media presence. Jeremy’s advice was to be deliberate rather than opportunistic. Decide what you want to talk about, where you want to contribute and how regularly you’re going to do it.
And then repeat the original exercise.
Go back to the same AI tools and ask the same questions again. Has anything changed? Are you appearing more often? Is the information more accurate?
Three takeaways from Jeremy's talk
If there were three ideas to remember, Jeremy brought them down to these:
- You can’t write your own reference.
What other credible sources say about your business increasingly matters alongside what you publish yourself. - Being absent is bad. Being wrong is worse.
Find out what AI tools are already telling people about your business before a potential customer does. - South Africa’s relatively thin content landscape is an opportunity.
There is still relatively little content standing between many South African businesses and the answer. But that opportunity won’t necessarily stay open forever.
The fundamentals of good communication haven’t disappeared. Businesses still need something useful to say, a clear point of view and the credibility to back it up.
What’s changing is where people go looking for that information.
For businesses thinking about their next phase of digital visibility, the question is no longer only “Can people find us on Google?”
It’s also becoming: “When someone asks AI about our industry, are we part of the answer?”
