From Amanda
WWDC 2026: Apple's AI strategy taking shape
WWDC 2026 felt like the first time Apple meaningfully joined the conversation about the future of AI.
Across Siri, App Store discovery, retention, and subscriptions, Apple is embracing AI as a core part of how users discover, engage with, and stay subscribed to apps. And in many cases, it’s using AI to address the abundance problem AI created in the first place.
Building apps has never been easier — releases were up 60% year-over-year globally in Q1 2026 — and subscription fatigue and discoverability are getting harder as a result (something I dug into recently with Steve Young). Nearly every major announcement from Apple this year seems aimed at helping users navigating that crowded space: Personalized Collections and App Notes surface apps based on behavior rather than keyword rank. And Retention Messaging and subscription bundles give developers new tools to improve retention and distribution.
And after years of promises for Siri, it seems to finally be catching up. App Intents is now the mandatory Siri integration surface. Rather than executing a single command, Siri can now understand goals, chain actions together, and complete tasks across apps without the user ever opening them. Apple is also asking developers to apply semantic metadata to individual UI components so that LLMs can understand what's on screen and act on it. For developers, this means exposing significantly more of their apps through App Intents. As for its impact on mobile growth, we'll be watching closely to see how a more agentic Siri intersects with traditional flows.
What's notable is that Apple's vision of AI looks very different from what we're seeing elsewhere in the industry. As Ben Thompson pointed out, Apple appears less focused on replacing apps with AI and more focused on using AI to help users navigate across them. Siri's advantage isn't the model, but that it understands your personal context and can operate across the apps already on your device.
On the measurement front, there were no major ATT updates or attribution shakeups this year. Privacy-first measurement is no longer the headline. At the heart of this year's announcements was Apple's belief that apps will remain central to the user experience, even as AI reshapes how people discover, navigate, and engage with them.
AI
AI is making marketing easier and measurement harder
AdExchanger reported last week that marketers are getting more comfortable with AI across the ad stack, from creative generation to LLM-powered ad placements.
The headline is that AI makes marketing easier with more campaigns, more creative variations, more audience segments, and more optimizations. But what happens after all of that goes live?
Unfortunately for mobile marketers, measurement was already hard. Thanks to SKAdNetwork constraints, signal loss, platform black boxes, and privacy regulations, attribution was getting trickier long before AI entered the conversation. AI adds another layer: more signals, but not necessarily more clarity.
Channel-level AI can optimize within a platform, but it can't tell you what's driving incremental growth across your full marketing mix. The challenge has shifted from generating ideas and launching campaigns to understanding which ones are actually working — a question that’s only getting harder to answer.
One manifestation of this is AI search. Most marketers haven't considered how advertising on LLMs will affect their measurement methodology. Branch research shows consumers are already using LLMs and traditional search side-by-side. Meanwhile, as our very own Adam Landis recently put it, "LLMs use the internet like an open-book test. You're likely already seeing high-intent organic traffic referred by LLMs, but what happens when ChatGPT and Google claim the same purchase?"
That's exactly the kind of measurement complexity AI is introducing across the marketing stack. If you're thinking about how attribution will evolve as AI-powered discovery grows, Adam wrote an article with a few practical steps marketers can take to prepare.
Advertising
The hidden tax on mobile growth isn't what most teams think
Most mobile growth teams think their biggest UA problems come down to a bad channel, a weak creative, or a targeting miss, but the real challenges are less obvious.
Branch recently analyzed more than 105 million paid install claims and $90 million in ad spend and found five hidden “taxes” that quietly eat away at performance. None of them are particularly groundbreaking on their own. In fact, most teams encounter them every day.
A user clicks an ad and lands on mobile web instead of the app. An owned channel like email influences an install, but never receives credit. Two ad platforms claim the same install. A channel creates demand but gets no attribution because it wasn't the last touch. A campaign is still improving, but the budget runs out before the platform finishes learning. Individually, these feel like operational issues, but together, they add up to a significant and compounding measurement problem.
The good news is that none of these taxes require more budget to solve. They stem from measurement blind spots, workflow inefficiencies, and optimization decisions that pile up over time. According to the analysis, they can represent millions of dollars in recoverable efficiency on a $10 million UA budget.
In a year where everyone's looking for growth from AI, new channels, and new tactics, it's worth remembering that some of the biggest opportunities are hiding in the fundamentals.
Adam's Take
The AI advertising battle has begun and first blood goes to Google
We’ve entered the dawn of “conversational commerce” Ads are finally starting to appear in your LLM conversations, and this was never a surprise. Earlier this year — before the technology even existed — 87% of enterprise marketers expected to execute closed-loop transactions via AI search. And it’s easy to see why marketers are excited: LLMs deliver highly motivated traffic to advertisers. Shopify studies show customers referred by AI are 50% more likely to convert and spend an average of 14% more per order. LLM advertising is the next frontier.
Google has landed the first blow, and in my opinion is well positioned to stay in the lead. It currently commands around a quarter of the world’s digital ad market and has already started serving ads on Gemini, AI Mode, and above traditional search results. Therein lies the first of its advantages. Search “coffee maker” in AI mode today and you’ll get a sponsored link. Inspect the UTM and you’ll see something like . Translation: Google is deploying paid search ads on LLM results. It doesn’t need to build an AI advertising business from scratch — it already has one. It just needs to turn it on.
OpenAI usage is growing fast — 84% over six months — but Google is gaining market share and growing 9x faster. But that doesn’t mean OpenAI will take a beating laying down. After abandoning its Instant Checkout pilot due to weak advertiser adoption and low performance, the company has already launched an ad business in five countries, expects to make a couple of billion from it by year-end, and is forecasting over $100 billion in revenue within two and a half years. That’ll bring OpenAI to a third of Google’s current size. Impressive, If it can pull it off.
Meanwhile, the dark horse Amazon. After (finally) killing off the Rufus name, it claims its AI shopping assistant drove $12 billion of incremental sales from 300 million shoppers last year. Amazon’s Andy Jassy claims shopping history gives the company a distinct advantage, but as analyst Kaziukenas points out, Amazon “only know[s] those last few clicks before purchase. Horizontal agents [like ChatGPT] are learning about users by building memory from their conversations.” Whether contextual knowledge ultimately beats order history is still an open question.
The scoreboard so far: Google has growing market share and a headstart on a scaled ad business. OpenAI owns the bulk of the traffic today but it needs to build a scaled, effective ad business to win — something that took Facebook 6 years to accomplish. Amazon has purchase history, a massive vendor network, and delivery infrastructure, but still needs to crack discovery. Gloves are off, and the advertising battle for LLMs has begun.
Podcast
The Viral Growth Blueprint: How Nic Weber Turned Creators Into a Scalable Acquisition Engine
Creator marketing is evolving from a brand play into a serious growth channel. In the latest episode of How I Grew This, Nic Weber, co-founder and CPO of Noise, shares why creator-generated content often outperforms traditional advertising, how brands can scale creator programs efficiently, and what separates viral campaigns from expensive misses.
Webinar
Navigating the AI Discovery Shift
AI search is changing how consumers discover brands, but measurement hasn't caught up. In this webinar, we explore where traditional attribution models fall short, what marketers can actually measure today, and how leading teams are preparing for an AI-driven discovery future.