News → Agentic Marketing

Agentic Marketing | November 4, 2026

Indian ChatGPT Ads tests already fail the first-order math

Tanishq, Tata Neu, ixigo and Wakefit are in, but Tall Creative's Divyat Kanyal says Wakefit/AbhiBus-style first orders of Rs 500–1,800 cannot carry a ~Rs 2,000 CAC, and conversion is under 1.5% against ~3% on Meta/Google.

Why it matters: Keep India ChatGPT tests on a written CAC vs first-order kill rule; do not add the channel to the Q4 mix until a named residual beats that math.

Anthropic marketers ship the app; security is the MCP gate

On Humans of Martech 244, Alexey Komissarouk says Anthropic marketers vibe-code tools instead of filing tickets, then stall when an MCP connector turns the app into an attack surface. They built their own CDP — he'd still buy Census or Hightouch at a growth-stage startup — and treat the warehouse as the stack that actually matters.

Why it matters: Inventory every MCP or API connector on internal marketing tools before the next vibe-coded app goes live; the review is the product, not the build.

ChatGPT is 20% of Google only if you count web searches, not clicks

Adam Fry told Big Technology ChatGPT's search-like queries are about one-fifth of Google's, counting every question that hits the web. Ahrefs' Nov 3 pull of 118,671 sites still shows Google at 24.74% of tracked traffic and ChatGPT at 0.36%.

Why it matters: Never put "ChatGPT is 20% of search" in a channel plan without naming query-volume vs referral; the click number is the one that funds the work.

← All Agentic Marketing briefs RSS