Quick Takeaways
- 1Google paid Apple $20 billion in 2022 alone to be the default search engine inside Safari — confirmed in unsealed US antitrust court documents in May 2024.
- 2In January 2026, the direction reversed. Apple now pays Google around $1 billion a year for a custom 1.2 trillion parameter Gemini model to power the rebuilt Siri.
- 3That means the AI reasoning layer on approximately 1.5 billion active Apple devices — including a majority of Australian smartphones — is now Google's reasoning.
- 4When a customer asks Siri 'who is the best pool builder on the Gold Coast', it is a Google-trained model producing the answer.
- 5Traditional SEO is not dead — but AEO (answer engine optimisation) is now the primary lever for whether an Australian business is visible in AI answers.
- 6Four levers matter most: machine readability, local entity clarity, question-shaped content, and third-party authority signals.
- 7The businesses that establish those signals before their competitors do will build a ranking moat that compounds for years.
The moment the money flow reversed
For most of the past two decades, one line item quietly held up a huge portion of the tech industry — Google's payment to Apple to be the default search engine inside Safari. Safari is the browser that ships preinstalled on every iPhone, iPad and Mac. The exact figure only became public during the US Department of Justice's antitrust trial against Google. In May 2024, unsealed court documents confirmed the number — $20 billion, paid to Apple in 2022 alone, for one thing: to make Google the first thing a user reached for when they wanted to find something.
That is more than the annual revenue of most Australian industries. In 2020, Apple's SVP of Services Eddy Cue confirmed that Google's payment made up roughly 17.5% of Apple's entire operating income. It was money spent, quite literally, to protect the answer.
In August 2024, US District Judge Amit Mehta ruled that Google had operated an illegal monopoly in search. The Apple deal was singled out as one of the mechanisms that kept competitors locked out. The tech world spent the next year debating what should happen next.
Then in January 2026, without much fanfare, the direction of the payment reversed. Apple announced it would pay Google roughly $1 billion a year to license a custom 1.2 trillion parameter version of Google's Gemini AI model. The purpose was to rebuild Siri — the assistant that had been sitting on iPhones untouched for years — into "Siri AI", a full generative-AI assistant capable of answering, reasoning and recommending. The Gemini engine runs inside Apple's own data centres, but the intelligence itself is Google's.
For twenty years, Google paid Apple to protect access to the search box. Now Apple pays Google to protect access to the answer. If you run a service business in Brisbane, on the Gold Coast, in Helensvale or anywhere across Australia, this reversal matters more than most people are recognising. The customer's phone did not just get smarter. The customer's phone is now recommending to them, on Google's brain, whose business to trust.
What Apple actually bought
The custom Gemini model at the centre of the Apple deal is not a stripped-down mobile assistant. It is roughly eight times larger than Apple's own cloud AI models. It has been engineered by Google, tuned by Apple, and installed as the reasoning layer for iOS 26.4 and beyond. That release will reach an estimated 1.5 billion active Apple users worldwide — most of whom will never know the assistant has changed brains.
The technical arrangement is subtle but worth understanding. The Gemini model runs on Apple's servers, meaning Apple retains privacy control over the queries. But the intelligence — the connections, the training data, the ranking of authoritative sources, the shape of the answer — is Google's. Every time a user in Melbourne asks Siri "who is the best pool builder in the Gold Coast", it is a Google-trained model producing the shortlist. Every time a user asks Siri "what should I look for in a commercial cleaner", it is Google's reasoning producing the criteria.
That distinction matters. Traditional SEO taught us to compete for the top of a search results page. AEO — answer engine optimisation — teaches us to compete to be inside the answer itself. The Apple-Google deal collapses one of the biggest walls between those two worlds. Whether a customer types a query into Google, asks Gemini directly, or now speaks to Siri, the underlying reasoning is increasingly the same reasoning. The signals that make you visible to one make you visible to all three.
Why this matters for Aussie small business
Aussie small businesses have watched this shift build across the past 18 months, whether they realised it or not. In 2025, Gartner projected that traditional search volume would drop by 25% by 2026 as generative AI answers displaced clicks. That prediction is already visible in Australian analytics — we have seen client Google Search Console impressions rise while click-through rates fall, a pattern that only makes sense if the answer is being consumed on the results page rather than driving a click to the website.
The Apple deal accelerates all of that. Siri is used by an estimated 1.5 billion people each day. In Australia, iPhone penetration sits above 55% of the smartphone market. That means the majority of Australian consumers now carry a Google-brained assistant on their hip that can, and will, recommend businesses on their behalf — not as a search engine, but as a personal assistant. The difference matters because personal assistants are trusted differently. When a customer types "best boutique in Fortitude Valley", they scan five results and decide. When a customer asks Siri "which boutique in Fortitude Valley would you recommend", they hear one answer — and they believe it. That is a structural change in how buying decisions get made, and it is happening at the same time that most Australian small businesses have never heard the acronym AEO.
The four things that shift for AEO
The old playbook — rank in the top three organic results, capture the click, close the sale on your site — is not gone. But it is no longer the whole game. Four things now sit alongside it, and each of them is worth serious attention.
Siri AI, powered by Gemini, does not read your website the way a human reader does. It parses structured data — schema markup, entity signals, business information tagged in machine-readable formats. A business without proper schema is not just unranked. It is invisible to the reasoning layer of the assistant. The good news is that fixing schema is a one-off structural project, not an ongoing content grind.
Google's Gemini favours businesses with consistent, unambiguous identity across the web. Same business name, same address, same phone, same categories — everywhere. A pool builder listed as 'GCC Pool Solutions' on Google Business Profile, 'GCC Pools' on the website, and 'Gold Coast Pools' in a directory looks like three different businesses to the AI, not one. Fixing this is unglamorous and enormously high-leverage.
Siri is a spoken interface first. That changes how content should be structured. Headings written like real questions — 'How much does it cost to repair a pool pump in Brisbane?' — outperform corporate headings like 'Pump Repair Pricing' every time. This is not a copywriting trick. It is a matching game between the question the customer asks and the phrasing on your page.
AI models are not neutral rankers. They are trained on a corpus of trusted sources. Businesses mentioned in third-party media, referenced on high-authority sites, cited in industry publications, or reviewed by real customers gain trust that is genuinely difficult for a competitor to counterfeit. This is what we call a ranking moat — a durable advantage that compounds every quarter.
What to do this quarter — the six-step playbook
The good news is that AEO is not more expensive than SEO. It is simply differently ordered. Here is what the priority list should look like for an Australian small business between now and the end of 2026.
Audit your machine readability first. Run your site through a schema validator. Confirm your organisation schema is complete. Add FAQ schema on any page that answers customer questions. If you are unsure where to start, we run a free AEO audit at infinity1.com.au/aeoquiz that scores your site out of 100 in about a minute.
Reconcile your business identity everywhere. Google Business Profile, website, Facebook, LinkedIn, industry directories. Same name. Same address. Same phone. Same primary category. Repeat for every location if you operate across multiple suburbs.
Rewrite your top ten pages with real questions as headings. Not 'Our services.' Not 'Pricing.' Instead, 'How much does a full kitchen renovation cost in Brisbane in 2026?' — the actual question a customer would ask.
Build one authoritative third-party mention per quarter. A guest article, a podcast interview, a local news feature. AI models weigh the corpus of sources, and Aussie SMBs are massively under-represented in that corpus. Even one strong mention per quarter compounds fast.
Update your top pages every three to six months. Freshness is a genuine signal to AI reasoning models. Not a full rewrite — a revision. A new stat, a new case study, an updated year in the heading.
Register your business with the AI ecosystem where it counts. Google Business Profile, Apple Business Connect, Yelp for Business, industry directories. These are the touchpoints that feed the answer models directly.
The bigger picture
The reversal of the Apple-Google money flow is the clearest signal yet that the era of "search" as a distinct activity is quietly winding down. What is replacing it is a set of answer engines, each competing for the trust of the person on the other end of the assistant. Google has secured a decade of that trust already. Apple has now paid to install Google's intelligence inside the most personal device most Australians own. For Aussie business owners, the practical takeaway is straightforward. If your business is not visible to Google's reasoning layer in structured, machine-readable, geographically-clear ways, you are not going to be recommended by Siri. Not because Siri is punishing you. Because Gemini cannot see you.
Let's revisit what we've covered
- →Google paid Apple $20 billion a year for two decades to protect the search box.
- →In January 2026, Apple began paying Google $1 billion a year to power Siri with a custom Gemini model.
- →The reasoning behind AI answers on 1.5 billion Apple devices is now Google's reasoning.
- →Aussie SMBs need to shift focus toward AEO — machine readability, entity clarity, question-shaped content, and authority signals.
- →The window to establish those signals before the majority of your competitors do is open, but not indefinitely.
At Infinity1, we work with Australian service businesses to audit exactly where they sit on the AI visibility spectrum, and to build the compounding lead engine that keeps them visible as the models keep evolving. If you would like to see where your business currently ranks in AI-generated answers, book a free strategy call and we will map it out with you. The sooner you start, the sooner the compounding begins.
Frequently Asked Questions
Is Google paying Apple, or is Apple paying Google?
Both — for different services. Google has paid Apple approximately $20 billion a year since the early 2000s to be the default search engine in Safari. That relationship remains in place. Separately, in January 2026, Apple agreed to pay Google approximately $1 billion a year for a custom Gemini AI model that powers the rebuilt Siri.
Does this mean SEO is dead?
No. SEO and AEO share the same underlying signals — quality content, authority, technical health — but AEO also requires machine-readable structure and question-shaped content. Businesses that do both will outperform businesses that do neither.
What is the single most important AEO fix for a small business?
Schema markup. Without structured data, the AI cannot reliably identify what your business is, where it operates, or what it offers. Everything else builds on that foundation.
How does an Australian business get mentioned in AI answers?
Consistent local entity information across Google Business Profile, your website and industry directories — combined with third-party mentions on trusted Australian publications. AI reasoning models cite the same sources humans would trust.
Sources
- CNBC, "Apple picks Google's Gemini to run AI-powered Siri" (12 January 2026)
- TechCrunch, "Google's Gemini to power Apple's AI features like Siri" (12 January 2026)
- Marketing Dive, "Apple taps Google Gemini to power AI features in multiyear deal"
- US District Court for the District of Columbia, United States v. Google LLC, Memorandum Opinion, Judge Amit Mehta (5 August 2024)
- Gartner, "Predicts 2025: Search Marketing"
Explore Infinity 1 services:
Want Infinity 1 to help your business?
We specialise in AI-driven marketing for tradespeople and eCommerce businesses. Get in touch for a free consultation.
Contact Infinity 1