Apple Pays Google $1B to Run Siri
- Greg McConnell
- Jun 9
- 4 min read
The headline is that Apple caught up. The real story: Gemini just became the default AI on a billion iPhones, and distribution, not capability, won.
At WWDC this week, Apple confirmed it will pay Google about $1 billion a year for a custom 1.2-trillion-parameter Gemini model to run Siri. Read past the Apple drama: Google just became the default intelligence on more than a billion iPhones.
THE LEAD
On Monday, Apple stood on the WWDC stage and, in effect, admitted it lost the model race. The rebuilt Siri will run on a 1.2-trillion-parameter Gemini model built by Google, for a reported $1 billion a year. The press called it “Apple finally caught up.” That framing misses the marketing story entirely. Apple did not buy a feature; it handed Google the assistant layer on the most valuable consumer base on earth. Gemini already anchors Android. Now it is the default cloud brain on more than a billion iPhones too. The interface your customer talks to, the one that reads their screen and acts inside their apps, is increasingly one company's model on both platforms. The prize was never capability. It was distribution, and Google just won the surface. This week, find out what Gemini says when a customer asks it to recommend a product in your category.
THE TAKE
The consensus question in every marketing org right now is which model is best: Gemini, Claude, or whatever shipped last week. Stop asking. Apple just answered it for you by renting instead of building. When the most capable hardware company on earth decides a custom Gemini is cheaper and better than its own decade-long effort, the lesson is that frontier capability has commoditized and converged. What has not commoditized is distribution. The model that sits in the default slot, on the lock screen, in the assistant your customer already uses, wins the interaction before a single prompt is typed. Marketers spent two years auditioning chatbots. The chatbot was never the point. The default slot was, and this week it consolidated. Build for the assistant your customer actually has, not the one with the best benchmark.
Marketers spent two years auditioning chatbots. The chatbot was never the point. The default slot was.
ALSO WORTH KNOWING
With Gemini now the default on both Android and a billion-plus iPhones, one model sits between brands and nearly every smartphone consumer. That is a distribution position no ad platform has ever held, and a single chokepoint regulators and rivals will fight over for years. For marketers, the practical takeaway is concentration risk: your discoverability increasingly depends on one company's assistant. Diversify how customers find you outside the assistant, and start measuring how often Gemini surfaces your brand unprompted.
Apple will tag AI-generated photo edits with hidden SynthID watermarks, the same provenance standard Google uses. Provenance signaling is quietly becoming default infrastructure, not a compliance afterthought. As assistants and platforms learn to read these tags, undisclosed AI brand creative becomes detectable at scale, and the FTC has already signaled it cares. Get ahead of it: adopt a content-provenance standard for your AI-assisted creative now, and write a one-line disclosure policy before a platform writes one for you.
UNDER THE RADAR
Buried in the WWDC fine print: the Gemini-powered Siri will be delayed on iPhone and iPad in the European Union and blocked entirely in China. The default assistant is not one global surface; it is a patchwork drawn by regulators. That fractures any “optimize for the assistant” strategy along national borders. In the EU, regulation slows the rollout; in China, your customer's default intelligence will be a domestic model you cannot test from headquarters. The lesson the trade press skipped: there is no single assistant to win. There are regional defaults, each with its own data, ranking, and rules. Map your top three markets by which assistant actually ships there, and assign a named owner to each before you set a global AI strategy that only works in America.
FRAMEWORKS & VOCABULARY
Share of Prompt, refined. Your old metric was share of search: where you rank on the page. Your new metric is Share of Prompt: how often the default assistant names, recommends, or links you when asked, measured per market because the default differs by country. The rule: you do not rank anymore, you get retrieved. Use it: “Our Share of Prompt on Gemini is 12 percent in the US and zero in the EU.”
THE PRACTITIONER MOVE
Run a Share of Prompt baseline this week. Write 10 buying-intent prompts a real customer would ask in your category. Run each through Gemini, now the default on iOS and Android, three times, and log whether your brand is named, recommended, or linked. Score out of 30. Under 10 means you are invisible at the new point of decision. Spend $1,500 and 10 hours to repeat it monthly across your top three markets, and bring the number to your next leadership review.
Think about this: your customer no longer chooses a model. Apple just chose one for a billion of them.
Forward this to the CMO still optimizing for Google's search results instead of Google's assistant.
About mktg.ai
mktg.ai is the Creative Intelligence System for modern marketing. The platform unifies every creative asset, channel, and KPI in one place, connecting creative performance to spend in real time so teams can act at the layer consumers actually experience. Features include Ask mktg.ai for natural-language queries against your marketing data, and Daily Alerts AI for automatic anomaly detection that surfaces performance issues without waiting for weekly reports. On a $5M media budget, mktg.ai customers typically recover over $100,000 by reallocating 15 to 20 percent of spend toward higher-ROI creative within months. Learn more at mktg.ai.
SOURCES
Comments