Author: Greg Kihlström

Gartner reports inference spending passed training spending in 2026. For CMOs, that makes verification a recurring operating cost on someone else’s meter. Here is how to price it and which instruments to own.

Point-of-purchase evidence has a shelf life. PubMatic, Makeable, Google, and METR each pushed AI verification from a buying decision into a continuously running function. What CMOs should own, instrument, and write into contracts.

Gartner, IAB, Cloudflare, and a candid APAC agency panel all landed in the same week on one question: what evidence quality does a number need before a CMO can spend against it. A practitioner’s read on grading AI measurement.

EU AI Act Article 50 transparency obligations took effect August 2. Greg Kihlström on what changes for marketing operations, agency contracts, and brand-configured agents.

New research puts a measurable ceiling on multi-agent coordination: collective accuracy peaks near 16 agents and declines after. Why verification cost, not compute, now governs how far marketing teams can scale agentic execution.

Enterprise marketing reports 70% AI deployment and 85% missed launch dates. New research from Knak, StackAdapt, Algolia, and Progress Software shows AI capacity landed on drafting while approvals, assembly, and coordination still set the calendar.

This article was originally published on MarTech by Greg Kihlström. I sit through a lot of vendor pitches and conference keynotes. After enough of them, the slides blur. AI-powered routing. AI-powered insights. AI-powered content. By the third talk, you can’t tell whether you’re being sold the same product four…

Human sign-off on high-risk AI agent actions fell from 40% to 25% in six months even as fully agentic campaigns began shipping end to end. The July 25 Martech Futurist argues judgment scales by designing oversight, not by adding reviewers, with data from JumpCloud, BCG, Broadsign, and Microsoft.

The Brand Visibility for Agentic Commerce framework scores a brand’s data surface — what an agent can resolve, compare, and trust — and the implicit assumption is that the agent is a fixed quantity the data is measured against. The evidence says the agent is a variable too. A…

From Greg Kihlström’s Martech Futurist blog. For two years the marketing conversation treated the model as the prize. Buy access to the best one, wire it into a workflow, wait for the returns. The research surfacing now moves the argument somewhere less comfortable. The model has become the assumed…