Recently, I attended the Sydney SEO Conference 2026 organised by Prosperity Media. It was a packed day of sessions focused on where search is actually heading.

AI is changing how content gets found, how credibility gets built, and how brands need to structure what they publish, and the conference invited some of SEO’s most respected experts to share their takes, insights and case studies behind what works for 2026 and beyond.

For this blog, I’ll cover three of the key themes addressed in the event that matter most to GEO, AI SEO, digital PR, and content strategy, and how they relate to the future of search.

Note that I’m only covering a short snapshot of the day, and this isn’t a substitute for the extra context, discussion, and nuance you get by attending the conference in person.

AI SEO: Going Beyond SERPs – Kevin Indig

Kevin Indig framed AI SEO around two distinct layers: the visibility layer and the trust stack. Most brands are focused on one and missing the other entirely.

Visibility is about how you show up in AI responses.

Kevin’s framework focused on four levers:

  1. Additive (original insights): optimise for information gain and address missing questions, rather than repeating what is already on the web.
  2. Webutation (presence across the web): web mentions have a big impact on AI visibility, while authority still comes back to high-quality links.
  3. Freshness (how up-to-date the content is): content under three months old performs best.
  4. Style (AI-friendly writing): definitive writing matters, and entities help ground answers in facts, which means using exact statistics, names, and specific details.

The main takeaway here was that visibility is no longer just about ranking pages. It is about publishing content that adds something new, stays current, and uses definitive writing with exact facts, names, and statistics that help LLMs ground answers properly.

Trust is about whether AI sees your brand as credible once you appear.

That stack includes

  1. Evidence (proven outcomes): customer stories, original data, patents, and certificates all strengthen trust.
  2. Fact density: specific, information-rich content gives AI more to work with.
  3. Reputation (authority and status): positive brand perception supports credibility.
  4. Experts (credible sources): validation from high-profile experts, trusted companies, or recognised endorsements matters because people follow people.
  5. Validation (external approval): reviews, press releases, and customer testimonials all back up your claims.

This framework directly applies to GEO (generative engine optimsation) because AI systems do not just need to find your content, they also need enough trust signals to choose, cite, and rely on it in generated answers.

Digital PR – James Norquay’s Digital PR and GEO growth session and a digital PR panel with journalists

Digital PR was a hot topic during the event, and there were two sessions dedicated for it that provided key insights:

  1. James Norquay’s session on digital PR techniques and insights from what his agency has done and
  2. A separate open panel with journalists to discuss digital PR from a separate POV (the publisher/journalist’s side).

James Norquay’s session made the main point clear: digital PR is the key driver for LLM success at scale.

Two campaign types were covered. Quick win campaigns and Hero campaigns.

  • Quick win campaigns: React to trends, seasonality, and timely data to provide relevant insights quickly.
  • Hero campaigns: Use in-depth original research and then pitch it to journalists for broader, longer-term coverage. James also pointed to internal data, FOI requests, trending topics, and visual data journalism as strong inputs for campaign ideas.

A separate session that followed-up this topic later during the day that was key was that the Sydney SEO Conference covers was what journalists actually think about the practice of digital PR for SEO and GEO, what their opinion on it was and more valuably, how to get journalists to actually pick up your campaigns.

In their words:

  1. Tailor it: journalists can spot generic outreach immediately, so the pitch needs to match what they cover and what their publication is interested in.
  2. Give them everything upfront: what gets noticed is an email that clearly lays out the expert, the data, the insight, the credibility, and the case study.
  3. Make it easy to read: journalists are busy, so the best pitches arrive already packaged as a story, not as work for them to figure out.

How LLMs work and how it impacts your content strategy – Frank Duignan

Frank Duignan’s session focused on how to optimise content for GEO, with three core areas: content chunking, natural language processing, and query fan-out.

Content chunking

LLMs break text into smaller sections, with a standard chunk sitting at roughly 512 tokens or about 400 words. You are not competing against other pages, but against other chunks. When retrieving text, LLMs prefer chunks of around 400 words, but page-level context still matters because of parent document retrieval. That means each chunk should be self-contained, sit within the right context, and create opportunities for LLM personalisation.

Natural language processing

To interpret human language, LLMs apply numerical values to words and draw on knowledge graphs to support that process. Frank highlighted semantic triples and statistics addition as practical ways to make content easier for AI systems to interpret, connect, and use more accurately in generated responses.

Query fan-out

Large language models run multiple simultaneous searches to build more detailed responses, draw on an online consensus, and personalise outputs for users. The practical takeaway was that your content strategy needs to reflect fan-out queries, not just the initial prompt, or you leave useful ground open for better-structured competitors to cover.

For GEO, this matters because content now needs to be structured for retrieval, interpretation, and expansion, not just page-level rankings.

 

Underlying theme for 2026: best practice is now common practice

One of the clearest ideas running through the day was that best practice is no longer enough on its own. As search changes, the baseline is getting higher, and more of what used to create an edge is becoming easier to replicate.

That point came through in different ways across the sessions. One of the strongest examples was the discussion around how many best-practice tasks can now be automated using tools like Claude. When strong execution becomes easier to scale, it also becomes easier for more brands to reach the same standard. That is what makes best practice increasingly common practice.

The takeaway is not that best practice no longer matters. It still does. But it is no longer where the real advantage comes from. What creates separation now is the layer above that: stronger thinking, better judgement, clearer positioning, more original insights (additive content), and a brand people can trust (built through authority building or digital PR).

This article only captures a snapshot of those ideas (only 3 out of the 10+ sessions). What made the Sydney SEO Conference valuable was the extra context, nuance, and live discussion behind them. If you want a clearer sense of where search is heading, it is well worth attending in 2027.

You can visit https://www.tickettailor.com/events/prosperitymediagroupptyltd/2116183 to book a seat at their next event in 2027.