From Core Updates To Answer Engines: What Actually Changed In 2025

The headline is simple. The Google core update 2025 cycle, especially June and September, was not just another quality refresh. It was a structural reweighting toward content that can safely feed AI Overviews and other generative features. To see why, you need to look at three converging shifts. First, Google is under pressure. OpenAI’s ChatGPT, Anthropic’s Claude, and tools like Perplexity have trained users to expect direct, conversational answers. The World Economic Forum and Stanford HAI have both highlighted how generative AI is changing information seeking behaviour, especially for complex decisions. Google’s response is AI Overviews, which sit above traditional results and synthesise answers from multiple sources. Second, regulators and media scrutiny are intense. After the messy rollout of earlier AI experiments, Google needs to minimise hallucinations, copyright blowback and safety issues. That means the June 2025 core update and the September 2025 Google update both leaned harder into trust, provenance and “information gain”, a concept Google has been signalling since 2022. Pages that simply rephrase what is already ranking are now a liability, not an asset. Third, user behaviour has shifted. According to Deloitte Digital and Think with Google, people are using more conversational, multi step queries. They expect to compare, simulate and decide, not just skim. AI Overviews are tuned to these behaviours, which means the underlying ranking systems are too. This means the old SEO playbook, which relied on: • chasing individual keywords • publishing thin “best X in 2025” listicles • buying or trading backlinks at scale • spinning generic “how to” content is now actively misaligned with how Google and AI assistants evaluate usefulness. The 2025 core updates are effectively telling you: "If your page would embarrass us as a quoted source in an AI Overview, it will not perform well." The deeper shift is from Search Engine Optimisation to Answer Engine Optimisation. AEO asks a different set of questions: • What are the real decisions, anxieties and tasks behind this query? • If an AI assistant had to answer in two paragraphs, what would it need from us? • How do we show evidence, nuance and trade offs, not just keywords? • How do we structure content so a model can safely quote, summarise and attribute it? Consultancies like McKinsey & Company and BCG are already retooling their thought leadership around these questions. Smart brands are following. The rest are watching their impressions rise while clicks and conversions quietly erode.

10 Ways The June And September 2025 Google Updates Killed Old-School SEO

To make this concrete, here are ten specific ways the June 2025 core update, the September 2025 Google update, and AI Overviews have changed the game. 1. Keyword matching lost power, intent depth gained it Pages that lightly sprinkle “google core update 2025” or “AI overviews SEO impact” without offering real analysis are being filtered out. Google’s systems, described in various patents and blog posts, are now better at inferring whether a page actually resolves the underlying task, such as “how should a CMO respond to AI Overviews cannibalising branded clicks”. 2. Thin affiliate and aggregator content is collapsing The updates hit “best tools” and “top X” roundups that add no original evaluation. If your monetisation model depends on sending users to someone else with minimal value add, you are on borrowed time. AI Overviews can already synthesise these lists from higher trust sources like Gartner, Forrester and Statista. 3. E-E-A-T is no longer a checkbox, it is a survival trait Experience, Expertise, Authoritativeness and Trustworthiness are not new. What changed is the weight. Pages that show real-world practice, such as case studies, methodology detail, or data from ABS or CSIRO, are more likely to be surfaced and quoted. Anonymous, generic content is being sidelined. 4. Over-optimised “SEO text” is a negative signal The old habit of stuffing every variant of “september 2025 google update” into headings and paragraphs now reads as spammy. Google’s quality systems and AI Overviews both favour natural, concise language that mirrors how humans actually ask and answer. 5. Clickbait titles without substance are being punished If your headline promises “Everything you need to know about the Google June 2025 core update” but delivers 600 words of recycled news, expect volatility. Google is increasingly measuring “information gain” and user satisfaction signals. Shallow content that overclaims is at risk. 6. Brand strength and offline signals matter more As AI Overviews quote sources, recognisable brands and institutions gain an advantage. If a model can choose between a random blog and a known entity like Harvard Business Review, it will often choose the latter. This is where brand, PR and SEO finally converge. 7. UX and cognitive load are now ranking levers The updates favour pages that reduce friction. Clear headings, short paragraphs, meaningful subtopics and helpful visuals make it easier for both humans and models to parse your content. Confusing layouts, intrusive ads and walls of text are being discounted. 8. Entity understanding beats raw backlinks The 2025 updates leaned further into entity based understanding. Google is mapping people, organisations, concepts and relationships. Being clearly associated with relevant entities, such as “AI Overviews”, “Answer Engine Optimisation” or “Bushnote”, often matters more than raw link count. 9. Freshness must be real, not cosmetic Simply updating the year in your title will not save you. The core updates are better at detecting whether content has substantively changed. For fast moving topics like “google core update 2025” or “AI overviews SEO impact”, you need real updates, new data and revised recommendations. 10. Cross channel consistency influences trust If your website says one thing, your LinkedIn another, and your media appearances a third, models will struggle to form a coherent picture of your expertise. Consistency across owned and earned channels is becoming a quiet ranking and quoting advantage. In short, the 2025 updates punished content that treats SEO as a game, and rewarded content that treats it as a public service. The more your page reads like something a careful analyst at Deloitte or PwC would be comfortable citing, the safer you are.

AI Overviews And The New Economics Of Clicks, Trust And Visibility

The most visible change for many brands is not rankings, it is click behaviour. You might still “rank” number one, yet see traffic fall because AI Overviews answer the core question before users ever scroll. To respond intelligently, you need to understand three layers of impact. Layer 1: Zero click answers at the top of the funnel For simple informational queries, AI Overviews often satisfy the user. This is similar to featured snippets, but broader. If your strategy depends on monetising raw pageviews from basic questions, you are exposed. According to Ipsos and Think with Google, people are comfortable getting quick facts directly in the interface, especially on mobile. Layer 2: Assisted exploration for complex decisions For higher stakes decisions, such as “how should a government agency govern AI search use” or “which consultancy can help us respond to the September 2025 Google update”, AI Overviews act as a guide, not an endpoint. They summarise options, surface brands and frame trade offs. Being cited here is powerful, but only if the click through leads to depth and clarity. Layer 3: Training data for other answer engines Your content is not only feeding Google. It is also being ingested, directly or indirectly, by models behind ChatGPT, Gemini, Claude and others. If you become a trusted, quotable source on a topic, you gain visibility across multiple answer engines, not just Google. This means you need to reframe success metrics. Instead of asking, “How many clicks did we get from ‘google june 2025 core update’,” ask: • Are we being cited or summarised in AI Overviews and AI assistants? • When users do click, do they stay, scroll and convert? • Does our content shape the narrative that models repeat about our category? Behaviourally, AI Overviews compress cognitive load for users. They reduce the effort needed to form an initial mental model. Your job is to become the source that deepens that model, not the one that repeats it. For example, if AI Overviews summarise that “Answer Engine Optimisation is about structuring content for AI assistants,” your page should go further. It should show frameworks, decision trees, and real examples of AEO in action. This is where specialist partners like Bushnote, who combine [AI search optimisation](https://www.bushnote.com/ai-search-optimisation) with [strategy and campaigns](https://www.bushnote.com/strategy-and-campaigns), can help you design content that both earns citations and drives outcomes. The economics of SEO in 2025 are therefore less about raw traffic and more about: • owning key concepts in your category • influencing how AI systems describe your space • converting fewer, better informed visitors If you optimise for those, AI Overviews become an amplifier, not a threat.

From SEO To AEO: A Practical Playbook For 2025 And Beyond

Answer Engine Optimisation sounds abstract, but it is simply SEO updated for a world where machines summarise you before humans see you. Here is how to operationalise it without drowning your team. 1. Start with decision journeys, not keywords Map the real journeys behind your primary terms. For “google core update 2025”, a CMO might move through: panic, diagnosis, prioritisation, budgeting, vendor selection, measurement. Each stage has different questions and emotional states. Design content to meet those, not just the surface query. 2. Build “model ready” flagship pieces Create a small number of deep, evergreen assets that you want AI systems to quote. For example, a comprehensive guide to “AI Overviews SEO impact for Australian brands”, or a decision framework for “choosing an AI search optimisation partner”. These should include: • clear definitions and frameworks • original data or synthesis from trusted sources like OECD, ABS or CSIRO • explicit, well structured answers to common questions Think of these as your category manifestos. They are the pages you want Gemini or ChatGPT to find when they look for authoritative explanations. 3. Use structure without sounding robotic Models and ranking systems love structure. Use clear headings, short sections, and explicit answers, but keep the tone human. Avoid robotic repetition of “september 2025 google update” in every subheading. Instead, use natural anchor phrases like “In short”, “This means”, and “According to” to guide both readers and models. 4. Make authorship and expertise visible Tie content to real people. Include bios, credentials, and where relevant, links to talks, research or media appearances. If your head of digital has spoken about AI Overviews at an industry event, mention it. This helps Google’s systems and AI models connect your content to real-world expertise. 5. Strengthen brand and narrative, not just pages Because AI Overviews and other answer engines often prefer recognisable entities, brand building is now a ranking strategy. Your narrative needs to be consistent across your website, social channels and media. Services like [brand and narrative](https://www.bushnote.com/brand-and-narrative) and [digital marketing](https://www.bushnote.com/digital-marketing) from Bushnote exist precisely to align these layers so that when models look for “who leads in AEO for complex organisations”, your brand fits the pattern. 6. Measure influence, not just traffic Update your reporting. Track: • branded and concept queries, not just generic keywords • mentions and citations in AI tools where possible • engagement depth, such as scroll, time on page and downstream actions This reframes SEO from “how many visits” to “how much narrative power”. 7. Iterate with real experiments Finally, treat AEO as an experimentation program. Test different content formats, depth levels and structures. Use tools like Google Search Console, analytics platforms and even manual checks in AI assistants to see how your content is being surfaced and paraphrased. Adjust accordingly. In short, AEO is not a new religion. It is simply SEO that accepts reality: most users will first meet your ideas through an AI summary, not your homepage. Your job is to make that summary accurate, compelling and biased in your favour.