The daily briefing for AI marketers and operators.

TL;DR — TikTok opened its ads platform to AI agents via MCP server and developer toolkit, enabling agents to plan, launch, and optimize campaigns without manual intervention. The platform lets AI handle creatives, bids, budget shifts, and targeting—work usually done by media buyers at a dashboard. Tactical implication: request MCP server access this week; test one campaign vertical (seasonal promo, new SKU) with AI-driven optimization—measure setup time reduction, bid efficiency, and ROAS vs. manual buying over 30 days.

TL;DR — Claude Sonnet 5 became Anthropic's default model on June 30 at $2/M input tokens and $10/M output tokens through August 31 (normally $3/$15). Early users (Cursor, Zapier) report agents stay on plan, follow conventions, and complete multi-step tasks end-to-end. The pitch: agentic capability is now the baseline at every price tier. Tactical implication: if running agents for campaign planning, creative generation, or audience analysis, cost-test Sonnet 5 vs. your current model—calculate token spend reduction and agent reliability improvement over 45 days.

TL;DR — Digiday surveyed 1,900 marketers: 86% adopted AI, but only 46% use agentic AI in workflows; 32% of leaders named GEO/AEO their top priority for 2026, with 12% of 2025 budgets allocated there. Strategy breakdown: 34% of companies add question-answer content to surface in AI search, while 28% optimize brand mentions and earned media. Signal: AI search visibility now rivals traditional SEO. Tactical implication: audit your top 50 organic keywords for AI search presence; prioritize 5 keyword topics where AI models misrepresent your brand or omit you—create answer-focused content and measure AI citation share weekly.

TL;DR — Google added AI Performance Insights and Conversational Attributes to Merchant Center, letting retailers understand brand performance across AI surfaces and optimize product data for conversational search. Conversational Attributes is a new product data field for natural, spoken queries. Tactical implication: retailers should audit product feeds for low Conversational Attribute coverage—add natural-language product descriptions and benefits (e.g., 'perfect for summer hikes' vs. 'lightweight hiking boot')—then measure impressions and clicks from AI search vs. baseline over 3 weeks.

TL;DR — Cloudflare announced AI crawler blocking is now the default for many websites, with new controls and publisher payment models. The shift means your content is blocked from AI training and search by default unless you explicitly opt-in. Brands lose AI visibility (SERP, generative search) but retain content ownership. Tactical implication: review your Cloudflare settings this week—if you want AI search traffic, whitelist Googlebot and other AI crawler IPs in your robots.txt or Cloudflare rules; measure AI search traffic and referral lift over 4 weeks vs. baseline.

TL;DR — Salesforce introduces immersive marketing as a way to engage prospects through interactive experiences. Tactic: Use virtual reality or interactive content to create memorable brand interactions. Play: Implement a pilot immersive marketing campaign to test engagement metrics this quarter.

TL;DR — MarTech highlights that AI visibility is significantly influenced by the authors covering your brand. Signal: Brands should focus on building relationships with influential writers in their industry. Play: Identify and engage with top content creators to enhance brand visibility through AI-focused narratives.

TL;DR — Yahoo DSP connected 23 ad tech partners into an Agent Network for agentic media buying, supporting audience targeting, creative, and measurement workflows without replacing existing tools. The network uses pre-built 'skills' and connectors to standardize workflows across agencies and DSPs. Tactical implication: test agentic workflows on one campaign vertical (e.g., seasonal sale)—measure setup time, manual input reduction, and ROAS lift vs. traditional buying over 30 days.

TL;DR — Time and The Economist are converting content from HTML to markdown, a format AI agents read more efficiently. Time strips navigation/styling and uses whitelisted bot access; The Economist is creating agent-readable versions of non-paywalled content (marketing copy, B2B material) to optimize for agent-driven discovery. Tactical implication: audit your high-traffic content pages—convert to markdown or simplified structure if you appear in AI search results; measure AI traffic and agent citations vs. baseline over 6 weeks.

TL;DR — Databricks launched CustomerLake, an agentic CDP that provisions autonomous agents for continuous customer analysis and real-time personalization at 1 billion decisions per day. Unlike legacy CDPs with week-long waterfall campaigns, CustomerLake runs agents natively on Databricks Foundry for instant context and activation. Tactical implication: if using a legacy CDP (Segment, mParticle, Treasure Data), pilot CustomerLake on one customer cohort—measure time-to-activation, personalization depth, and incrementality vs. your current tool over 45 days.

TL;DR — DeepIntent unveiled Helix AI, an agentic platform for healthcare marketers to query patient and provider data in natural language, build audiences, and activate via DSP and social channels. The platform addresses healthcare's dual-audience challenge (patients + medical professionals) with unified intelligence and activation. Tactical implication: healthcare marketers should request beta access—test on one audience segment (e.g., patients with diabetes diagnosis); measure audience precision and campaign setup time reduction vs. manual workflows over 30 days.