From Amanda
Apple: Gemini and protecting the privacy narrative
Apple has confirmed a multi-year partnership with Google to use Gemini AI models and cloud technology as the foundation for future “Apple Intelligence” features, including a revamped Siri expected later this year. This pivot follows an underwhelming AI debut at last year’s WWDC and a broader AI strategy that has visibly lagged competitors.
This isn’t the first time these “rivals” have teamed up. Google already pays Apple billions to stay the default search engine. But Gemini sits a lot closer to the core experience than search ever did. And as The Verge points out, Apple immediately wrapped the announcement in privacy language: Gemini may power the backend, but Apple positions it as running through Apple’s privacy architecture (on-device plus Private Cloud Compute). The message is pretty clear: we’re partnering, but we’re still in control.
Meanwhile, Apple’s privacy posture is also under fresh scrutiny. The Italian Competition Authority ruled that Apple’s App Tracking Transparency implementation unfairly harmed developers through double consent requirements, issuing a $115M fine. It’s no surprise that Apple strongly disagrees and plans to appeal while remaining steadfast in its privacy-forward messaging.
Apple’s AI changes extend beyond partnerships and messaging. The company is replacing longtime AI chief with the previous leader of Vision Pro—notable timing, given Vision Pro continues to fall short of expectations and Apple has been pulling back on manufacturing and marketing. Stratechery's Ben Thompson has argued that its struggles go beyond price or hardware, pointing instead to its weak product narrative and unclear long-term value.
Which brings us back to why Apple’s narrative matters so much right now. The company’s moves may feel reactive, but it’s still selling privacy and control.
AI
The rise of AI and automation will increase appetite for humanity
Eric Seufert’s 2026 mobile predictions reinforce how deeply AI is penetrating the marketing stack. Platforms like Meta are pushing toward this vision of full end-to-end automation with generative ad creative and AI-driven targeting and optimization.
But as Ben Thompson has said recently: AI scales compute, but humans scale connection. That feels like a useful lens for what’s happening with the big platforms.
The more ads become automated, the more it matters where an ad appears, who people trust there, and why they’re showing up. That’s driving renewed interest in environments that feel intentional and human, like advice forums, creator-driven ecosystems with distinct points of view, and editorial spaces people trust.
I expect we’ll see a growing appetite for humanity in practice this year — something Yanni Pappas touched on recently. It will be interesting to see how that shows up as AI becomes more embedded in everyday marketing workflows, and where advertisers decide to trust the machine. It’s almost counterintuitive: The more automated things become, the more we value what feels expressive, shaped by taste, and made with intention. In that sense, the AI boom might actually bring us closer to humanity rather than pulling us away from it.
Walled Gardens
Automation comes to a growing Reddit, but will it click?
Reddit just rolled out Max Campaigns — it’s version of Advantage+ and PMax, with AI handling targeting, creative, and delivery, all powered by the vast community data Reddit sits on.
The company is seemingly taking a page out of the major platforms’ playbook. Partnerships with Google and OpenAI have expanded Reddit’s reach through search and AI answer engines. It even overtook TikTok as the fourth-largest social platform by growth in the UK, and eMarketer shows Reddit’s mobile ad revenue climbing in the US.
But what really caught my attention was how Reddit chose to talk about performance. As Eric Seufert pointed out, the press release emphasizes clicks rather than ROAS.
Clicks on Reddit have always signaled trust: People engaging because they believe the advice came from real humans. Now, we’re seeing AI-generated ad creative and even AI-generated posts go viral, including a recent fabricated fraud claim (listen to the most recent Hard Fork episode for a deep dive).
How ad dollars follow this shift, and how users respond, will be worth watching.
Industry Buzz
What does the impending AppsFlyer acquisition tell us
Reports say AppsFlyer is in talks to be acquired by private equity. If it materializes, it’s less a headline about one company and more a signal about the enduring role attribution plays in the mobile stack. As measurement becomes more predictive and ad platforms automate more decision-making, advertisers still need an independent layer to collect signals, validate performance, and act as a trusted source of truth across channels.
Adam's Take
Mobile’s lesson for AI
Everyone knows AI is important for business but we’re struggling to understand the impact or how to adopt the technology. To me the answer is simple: We need to measure AI’s impact on our business, then iterate where it works. Mobile offers a lesson in how this will take shape.
We’re at late stage mobile adoption now. Most industry growth comes from replacement devices, not net-new subscribers. Having experienced the entire growth curve, I’m seeing patterns repeat in today’s world of AI.
In the early days of mobile, we'd sit on the App Store hitting refresh to track rank changes — that was our only growth indicator. Eventually the industry developed analytics for better tracking (e.g. mobile measurement partners like Branch), which allowed us to understand marketing impact. However, once Apple broke deterministic measurement by introducing SKAN, we learned to measure our marketing efforts through signal and probabilistic models.
Today it's early days in AI. Instead of refreshing the App Store, we’re asking the LLM to tell us where we rank (yes, there are even tools for that). AI systems are a black box, and we can’t get deterministic outputs. Given AI’s rapid, expansive saturation and complexity, there’s very little likelihood that deterministic measurement will ever be available, or effective.
But therein lies the lesson: Mobile taught us that the outcome of mature, complex multi-channel marketing is probabilistic, influence-based models. There is no single right reason for a conversion. Instead, marketers need to measure the combination of signals. Case in point: a recent study showed online sales represented 20% of all retail transactions. But when factoring influence, digital represents 60% of all retail transactions. To understand AI's influence as a channel, we can’t wait for a deterministic “last-touch” model. We need to understand the complete influence today.
The future winners in AI aren’t waiting around for perfect AI analytics — they’re investing in understanding real business impact today.
Podcast
The Brand Consistency Myth: How to Scale Creative Without Losing Your Identity with Justin Rashidi
In this episode of How I Grew This, host Amanda sits down with Justin Rashidi, Co-Founder and Head of Strategy at SeedX, Inc., alongside co-host Adam, to explore how to build sustainable growth engines, navigate the shifting search landscape powered by AI, and measure marketing efforts with precision in a world of constant platform changes.