Google, Meta and OpenAI Trends Marketers Should Watch
ChatGPT Ads Are Coming (maybe?)
Leaked documents show OpenAI is preparing to introduce ads directly inside ChatGPT conversations.
What's the News:
Unlike Google, ChatGPT operates on a lagging training dataset with limited real-time signals — a major constraint for any ad system. OpenAI has temporarily paused rollout to shore up model quality after Google’s Gemini 3 leap.
Why It Matters:
ChatGPT ads won’t behave like Google Search ads. Without a live, comprehensive index or decades of intent data, early ChatGPT ads will struggle with:
- interpreting time-sensitive or trending queries
- matching users to the most relevant advertisers
- expanding into broader-but-related intent with precision
Expect lower relevance, wider performance variability, and lower CPCs simply due to weaker targeting signals. There’s also real brand trust risk: users may reject sponsored answers inside a tool they view as a neutral assistant. Creative will need to feel native — more Quora/Reddit-style than search ads — while still clearly distinguished as sponsored.
ChatGPT ads will become a real channel, but brands who jump in early should be cautious. Performance will lag Google until OpenAI solves its real-time data deficit, and perception risk is high if ads feel intrusive or bias the model’s answers. Marketers should view early participation as experimental only, and start preparing creative designed for conversational, AI-native ad environments. Long term, getting listed through ChatGPT ads may prove more consistent than earning AI search visibility — but the model needs to catch up before the economics work.
Google Is Poised to Win the AI Race
What's the News:
New eMarketer data predicts Google will overtake OpenAI in monthly generative AI users by 2026. ChatGPT’s share of global genAI traffic has dropped from 86.6% → 72.3%, while Google’s Gemini has nearly tripled to 13.7%. Google already reaches billions through Search, Chrome, Android, Gmail, Maps, and YouTube — and blind tests show consumers prefer Gemini over ChatGPT for search, text, and image generation.
Why It Matters:
The center of gravity for AI-assisted discovery is shifting toward Google — not OpenAI. With 2B+ users seeing AI Overviews each month and 3B+ having AI Mode access, Google has the distribution, real-time data, and product ecosystem to become the default AI assistant. This changes how brands get discovered, how content is evaluated, and how consumers interact with information. Marketers who treat ChatGPT as the primary AI channel risk optimizing for the wrong platform.
Google is on track to own AI discovery the same way it owns search — and marketers should adjust now:
- Prioritize GEO (Generative Engine Optimization): Optimize content for AI Overviews and question-based retrieval, not just traditional rankings.
Diversify your AI toolset: Test Gemini alongside GPT models for creative, copy, and analysis to avoid OpenAI dependency. - Prepare for mobile-first AI interactions: As Gemini becomes the default assistant across Android and Chrome, brands need conversational, task-oriented content that converts inside AI flows.
- Assume Google, not OpenAI, will set the rules for visibility, attribution, and AI-era reach.
The market is signaling a clear pivot: stop operating with a ChatGPT-first mindset. Start building for an AI ecosystem where Google becomes the dominant entry point for discovery, answers, and action.
20 Is the New 50? Meta’s Learning Phase Might Be Changing.
What's the News:
Meta appears to be testing a reduced learning phase threshold — from 50 conversions down to 20. It’s not universal yet, but several accounts are showing “20/20” learning indicators, suggesting a partial rollout.
Why It Matters:
A lower learning requirement changes how quickly campaigns stabilize and how much budget you need to validate performance. Marketers can expect:
- Faster stabilization and less “learning tax.”
- Lower validation budgets — new rule of thumb: daily budget ≈ (CPA × 20) / 7.
- Better viability for high-CPA events like Purchases, Subs, or Qualified Leads.
- More room to optimize to the event you actually care about — not just upper-funnel proxies.
There are trade-offs:
- Noisier signals with smaller sample sizes.
- Consolidation still wins — 20 events is not permission to fragment your structure.
- Major edits still reset learning, just as before.
If your account is showing “20/20,” use it — but don’t build your entire strategy around it. Plan pacing and budgets to the 20-event threshold, but still push for 30–50 events to maintain stability. Keep your account tight, consolidated, and signal-rich (CAPI, deduping, high EMQ). Batch changes, give winners room to run, and avoid unnecessary resets.
TL;DR:
This shift speeds up ramps and lowers the cost of validation — but only if your structure and signal quality are strong. Curious who else is seeing the 20-event rollout and how it’s impacting stability.
We’re reinventing performance marketing because everything changed while you were reading this.
We help you navigate this new landscape with data-driven insights, platform expertise, and creative that connects, so your brand not only keeps up, but gets ahead.
Work with Us