The Behaviour That Changed Our Inbound

Classic SEO assumed people began on a page of links and then chose. That still happens. But many journeys now begin above the links, inside systems that compose an answer and attach a handful of sources. Google calls this layer AI Overviews. Perplexity describes a similar process plainly: it searches the internet in real time and distils what it finds into a concise, sourced answer.

Two market signals make this shift hard to ignore. First, independent clickstream studies find that for every 1,000 Google searches, roughly 360–374 clicks go to the open web; the rest stay on Google or end with no click. Second, adoption keeps climbing: a third of U.S. adults have used ChatGPT, including a majority of under‑30s. Australia trails by months, not years. If you sell into English‑speaking markets, this is already your distribution channel.

Intent, Rewritten: How Decisions Form Now

The old model neatly sorted behaviour into informational, navigational, commercial, transactional. Those buckets still exist; they’re just no longer where many journeys start. Two patterns now sit above them:

• Generative intent — people open with a broad question and accept a composed summary as the first draft of truth.
• Exploratory intent — people use the model as a thinking partner, test directions until a goal forms, then proceed with a shortlist.

This has practical consequences. Pages that resolve a real question, expose clean structure the model can parse, and seat proof next to claims are easier to include — and to quote accurately. Pages that bury the answer, waffle through adjectives, or hide receipts are easy to omit. It isn’t mystical; it’s how these products are designed: snapshot first, sources attached, user invited to verify.
‍

How We Stay in Front

We’re a specialist marketing agency. We don’t win on headcount or history. We win by keeping our research loop close to how these systems actually behave. In practice, that means three ongoing disciplines:

A. We monitor the models and the platforms.
We track how AI Overviews, ChatGPT and Perplexity are selecting and displaying sources, and what they appear to value: clarity of resolution, structural consistency, recency signals, and absence of conflicts. Preferences evolve; our approach evolves with them. This is not a launch moment. It is upkeep.

B. We engineer for people first, and for reuse by systems.
Humans get a clear promise and a resolved answer quickly. The same pages give models a predictable shape to read, with verifiable evidence beside the statements that matter. That combination reduces mis‑summaries and increases inclusion.

C. We maintain the asset, not the calendar.
We strengthen the few pages that shape decisions, add new proof as it is earned, retire what dilutes the signal, and keep source‑of‑truth pages current. It’s an operating model, not a campaign.

“In recent weeks, a material share of new leads began in ChatGPT rather than classic search. That validated the work we’ve invested for months — writing for people and for the systems that summarise for them, then maintaining those assets like infrastructure.” James Dore, Marketing Director, Bushnote

‍

The Implications

This isn’t a marketing curiosity. It moves risk, revenue and reputation across the economy — from sole traders to listed multinationals. Think of these patterns you can map onto your own world:

Reputation‑sensitive organisations.
If you operate in energy, healthcare, finance, higher education or government, summaries can freeze a past narrative if your current position is hard to find or parse. When models reach for sources, they may over‑weight conversational platforms or dated reporting; even Google warns that AI responses “may include mistakes.” The remedy isn’t spin; it’s maintained, machine‑legible source‑of‑truth pages an engine can cite and a journalist can check.

High‑consideration consumer purchases.
Mortgages, telco plans, private health, EVs. Buyers now begin with “Who should I talk to this week?” not “What is [category]?” If your expertise is easy to summarise accurately, you enter the shortlist with trust already forming. If aggregator content or generic definitions dominate, your team opens at price instead of value.

Procurement & B2B software.
Time‑poor teams ask models for a shortlist that satisfies technical, security and integration constraints. If your certifications, API coverage and customer evidence are scattered or out of date, the model fills gaps with third‑party summaries. You may be the right fit yet still be absent from the paragraph that books the meeting.

Listed companies & markets.
Disclosures move prices. Regulators have warned against “AI‑washing” — overstating AI capabilities — and exchanges expect precision. When a model compresses your announcement into two sentences, any ambiguity becomes a price risk. Keep the primary source accurate, current and easy to quote, and make sure the first lines say what you’d defend to a regulator or exchange.

‍

Field Notes from the Front Line

AI‑led leads arrive warmer. Prospects quote the answer they saw and move straight to constraints. The first conversation already happened in a summary.

Paste‑ability predicts spread. The lines stakeholders copy into internal emails are often the lines models reuse. Write so your key statements can travel intact.

Authority stacks when earned. Once you’re cited cleanly in one model, cross‑verification accelerates elsewhere.

Volume without stance reads as silence. Weekly posts that never take a position vanish from human memory and machine context alike.

“Perplexity has between a 6× and 10× higher click‑through rate than ChatGPT … and the conversions from Perplexity are excellent.” — Josh Blyskal, AEO Engineer, Profound (Siege Media podcast, Jun 30, 2025)

What Leaders Do Now

Treat this as infrastructure, not content. Decide the few questions that fund your year. Keep the pages that answer them alive, accurate and summariser‑friendly. Make sure the first 150 words say something useful, and that proof lives next to the claim it supports. Monitor how the major platforms are presenting and citing your material and adjust your source pages accordingly. It’s a continuous investment, not a one‑off push.

Are You In the Answer?

We’ll keep doing the quiet work that put a new name at the front of “Top SEO agency in Australia” conversations: watching the models, hardening our own sources, and helping organisations become easy to quote for the right reasons. If you lead a government team that needs accuracy at first read, a national brand that wants to be the definitive answer in its category, or a listed company that cannot afford mis‑summaries at market open, we should talk.

Before you close this tab, run one test. Open the tool your audience actually uses. Ask the three questions that matter. Read the paragraph before the links.

Are you in it?

If not, that’s the conversation we should have.

‍