Analyzing 50 000 LLM‑generated query fan‑outs, Moz found brand mentions appear in nearly 13 % of prompts, driven almost entirely by two fan‑out types, Entity and Comparison. Alignment scores (cosine similarity to subtopic) average 0.67, showing the prompts are relevant but not overly narrow. The raw dataset is freely downloadable.
A Fractl survey finds 82% of marketers have earmarked budget for AI search, with 24% of overall search or content spend already allocated and 43% spending over 20% of that bucket. SEO and performance teams lead the spend, yet 18% still invest nothing, highlighting a fragmented AI‑search infrastructure.
In a study of 1,000 unbranded legal questions, Google’s AI Overviews included at least one provider in 59% overall and 92% of commercial‑intent queries, with an average recommendation list of 7.6 firms. Educational content frames the recommendation, making content SEO a key lever for legal visibility.
X posted $367 million in advertising revenue for the April‑June quarter, a 66% plunge from the $1.08 billion Twitter earned in the same period two years ago. The drop makes ads a tiny slice of SpaceX’s $7.8 billion total revenue, shifting the company’s growth focus to AI and connectivity services.
Time magazine is running a hidden version of its site that only AI crawlers see, serving sponsored content from brands like Ally Bank. By feeding AI assistants tailored ads, publishers can shape the answers these bots give users, turning AI assistants into a new ad medium and forcing marketers to plan for machine audiences.
Nielsen announced an all‑cash $2.15 billion deal to buy DoubleVerify at $13.60 per share, a 30% premium. The merger adds independent ad‑verification and brand‑safety tools to Nielsen’s audience data, giving advertisers and publishers a single platform to measure, verify, and optimize media across every screen.
Agentic marketing uses AI agents that take a goal, like “triage today’s comments”, and decide the exact actions, tools, and timing themselves. Unlike traditional rule‑based automation that breaks when reality diverges, agents read results in real time, adjust, and keep the system alive, cutting the need for constant workflow rewrites.
You’re caught up.
That was all 11 stories in the 7 August issue.