Forget LLMs.txt: AI ads lift brand search 185%

Google’s John Mueller revealed that LLMs.txt was never intended to help search engines discover content, contradicting the way many sites are using it to boost AI visibility. The mismatch means effort spent on LLMs.txt for ranking is wasted, and the file’s self‑reported data is unreliable for crawlers.

A study of millions of AI chat interactions found that when an AI platform recommends a brand, users become 182% more likely to Google the brand, 117% more likely to visit its site, and 185% more likely to view its products within a week. The findings prove AI search, despite being a zero‑click channel, drives measurable downstream traffic and forces marketers to optimize brand visibility for AI answers.

Marketing leaders keep piling on courses, but the real issue is a lack of decision‑making practice. Marketers can’t explain their reasoning or sit with uncertainty, fueling imposter syndrome and poor stakeholder communication. Building judgment, not just knowledge, is the only way to close the gap.

New data from the Ehrenberg‑Bass Institute shows only ~10 % of B2B buyers actively reject a brand, while the average brand has 1.6 × more potential customers unaware of it, in small brands the gap widens to 8-16 ×. The takeaway: prioritize memory‑building over perception tweaking to drive sales.

OpenAI’s Ads Manager is now live in the United Kingdom, marking its fifth market rollout, and the platform adds cost‑per‑click (CPC) bidding alongside the original CPM model. The new options let UK advertisers align spend with clicks and tap a self‑serve portal to launch and track campaigns directly in ChatGPT.

Fanatics Betting & Gaming boosted projected customer LTV by 19% by ditching predefined audience segments and letting Cognitiv’s AI optimize CTV campaigns directly for value. The shift to outcome‑based media buying links ad spend to high‑value bettors, signaling a move toward machine‑learning‑driven audience creation across channels.

Penske Media Corp. bought Vox Media’s remaining assets, Eater, The Verge, SB Nation and PopSugar, and folded them into a new PMX division alongside its own titles like Billboard, Variety and Rolling Stone. The combined 25‑plus brands give PMC scale, hundreds of millions of monthly users and a live‑event and ad tech platform, forcing advertisers to rethink buying across a single inventory.

Survey by eClerx shows 78% of marketers say their martech investments don’t meet business goals, yet only a quarter feel fully data‑driven. The real barrier is operational, fragmented data, low confidence, and an “activation gap” that turns insights into inaction.

Building a Marketing Mix Model is futile unless you meet three design prerequisites: at least 18 months of consistent weekly spend and outcome data, enough variation in spend to detect effects, and paid channels that actually drive results. Miss any of these and your MMM will produce biased, misleading insights.