In partnership with
Thursday. Planning season, so let's do a planning question.
How many boxes on your 2027 org chart are not people?
Greg Kihlstrom's column in MarTech walks through four products converging on one idea. Optimizely's Virtual Teammates, Treasure AI's Marketing Super Agent, HubSpot's Breeze agents, Asana's AI Teammates.
The demos are not the interesting part. They all ship with the same three things: an identity, a set of permissions, an audit trail. That is not how you package a feature. That is how you onboard staff.
Then the numbers underneath, both Gartner's, both pointing away from the marketing. Over 40% of agentic AI projects will be canceled by the end of next year on cost, unclear returns and missing governance. And of the thousands of vendors selling agentic AI, roughly 130 are actually building agentic capability into their platforms.
130. Keep it in the drawer you open during demo season. Ask a vendor to show you the identity, the permission model and the audit trail, and you will know inside ten minutes which pile they are in.
The management point is the one I keep returning to. There is a difference between telling a system the twenty steps of a nurture sequence and telling it what you want and what it may not touch.
Kihlstrom's version of the second reads like a brief you would give a good hire: lift mid-market SQL conversion 15%, hold brand sentiment at 4.0 or better, never touch an account with an open Tier 1 ticket. Nobody on your team gets a twenty-step script. Why would the software.
His three screening questions beat most governance frameworks I have read this year. Is it reversible. How does the output get reviewed before it reaches a customer. What happens when two agents disagree.
That last one is where the org chart stops being a metaphor. Somebody owns drift, arbitration, sandboxing and the kill switch, and it has to be a named person.
So who on your team gets that job, and what comes off their plate the day you give it to them?
Today’s sponsor:
Stop losing deals in between meetings.
Aligned is changing how B2B teams run their most important deals.
Most reps guess where a deal really stands. They forecast on a gut feeling and the buyer’s word vs what's actually happening. Then they realize the deal slipped through only after it's too late. Aligned surfaces what buyers are actually doing inside every deal: the risks, the openings, and the exact next step to keep things moving.
Start increasing your close rates and stand up your first deal room in minutes.
TALK ON THE MARKET
Jed Fudally, director of demand generation at Siro, on pointing AI at his own pipeline data, in Dreamdata's launch release:
With generic AI, I'm confident it'll give me a response. I'm just not confident that the response is accurate. The Dreamdata Analytics Agent shows me exactly how the report was built, the filters, the model, the date range, so I can check it for myself. That's what earns my trust.
Vendor release, and he is their customer, and he has still named the real bar. Not how fast it answers. Whether you can see the working.
Want to see case studies of AI-powered campaigns?
Try our sister newsletter:
NUMBERS MATTER
13%. The fall in initial earnings for graduates in the most AI-exposed majors, with the likelihood of landing a first job down five percentage points, in a working paper by economists affiliated with the US Census Bureau, covered by Inside Higher Ed. Computer science, accounting, journalism and engineering all sit in the most-exposed bracket.
The paper's own comparison: "This earnings decline is comparable in magnitude to the earnings losses associated with graduating into a large recession."
Working paper, not peer reviewed. Also the rung your future senior hires are standing on.
DAILY PROMPT
Our Design Your Marketing Organization for the AI Era, which starts from the friction rather than the boxes:
You are an expert marketing operations strategist helping a CMO redesign their organization to embed AI effectively and reduce operational debt.
## Context
Most marketing teams are drowning in operational debt: coordination overhead, approval delays, tool sprawl, fuzzy ownership, and broken handoffs. This debt prevents AI from creating real value. AI tools plugged into broken workflows just hit the same bottlenecks faster.
The goal is not to "add AI everywhere." The goal is to rewire high-friction workflows where time is leaking and revenue is at stake, prove measurable lift, then scale.
Work through this as a multi-step analysis. Read every file I attach (org chart, tool and license inventory, workflow or SLA documentation, campaign calendar, CRM and pipeline exports). Run the numbers yourself where the data supports it. Research current governance and staffing patterns in comparable B2B software organizations if that sharpens the recommendation, and label anything you could not verify against my data as an assumption.
## Your Task
Analyze the marketing organization described below and design a lightweight, AI-ready structure that:
1. Eliminates the highest-friction operational bottlenecks
2. Creates clear ownership and decision rights for AI implementation
3. Builds in lightweight governance (security, brand, data risk) without killing velocity
4. Connects AI outputs directly to pipeline and revenue outcomes
5. Enables compounding value across pilots instead of siloed experiments
## Organization Details
**Current State:**
- Team size: [NUMBER] people across [DEPARTMENTS]
- Key pain points: [DESCRIBE 3-4 biggest operational bottlenecks: approval cycles, handoffs, tool sprawl, etc.]
- Current AI usage: [DESCRIBE existing AI tools, pilots, or shadow AI]
- Revenue impact: [DESCRIBE how marketing currently connects to pipeline/revenue]
- Governance status: [DESCRIBE current risk/compliance/brand guardrails]
- Constraints: [DESCRIBE budget, headcount, or platform constraints]
## Deliverable
Provide:
1. **Operational Debt Audit**: Rank the top 3 workflows causing the most time leakage and revenue friction, with the evidence from my files behind each ranking
2. **AI-Ready Structure**: Propose a lightweight org design with clear roles, decision rights, and accountability for AI implementation
3. **First 90-Day Roadmap**: Identify one high-friction workflow to rewire with AI first, with success metrics tied to pipeline/revenue
4. **Governance Framework**: Simple rules for brand, data, and security that enable speed, not kill it
5. **Scaling Plan**: How to compound value from the first pilot into the next 3 initiatives
Be specific about roles, decision-making authority, and how AI changes the work itself, not just the org boxes. Close with the three questions whose answers would most change your recommendation.Feed it the real org chart and tool inventory, not a description. The operational debt section is the one that earns its place: a team with a three-week approval chain does not get faster when you add an agent, it gets the same bottleneck more often.
CMO CORNER
Every task you take back is a future habit you reinforce.
True of people, and about to be true of software. The failure mode is identical: a system that never earns autonomy because somebody keeps stepping in at the last moment, then gets blamed for needing supervision.
Decide in advance what an agent is allowed to get wrong. If the answer is nothing, you have hired an intern and refused to let them near the work.
We provide consulting services for CMOs on AI strategy for marketing teams through our Teamless offering.
TOOL UPGRADE
Three ways to find out whether an AI coworker is a real idea for your team, at three very different prices:
Lindy — best for ops teams wanting multi-step workflows with real branching and no engineer, from $99 a month. Try Lindy here →
n8n — best if you have engineering help or need self-hosting, free to ten workflows, with a genuine learning curve. Try n8n here →
MarketerHire — best when the honest answer is a person this quarter rather than a workflow, from around $5,000 a month. Try MarketerHire here →
→ We have 300+ curated, filtered, and categorized AI tools in our directory. Browse to find the best ones that help your marketing team.
LEARN THIS
Build one before you buy five: Build Your 24/7 AI Agent. The point of the 35 minutes is not the build. It is coming out able to tell which demos are doing something and which are a form with a personality.
→ We publish workshops, courses, guides, manuals, and more on AI marketing for free. Browse our learning resources for free.
BRAND SLOP
Target apologized on 24 August and pulled a children's clown costume after critics said it evoked blackface. Then a video went round that looked exactly like a Target ad for it.
It wasn't. Reuters traced it to an X post the next day carrying a "Made with AI" label, built with OpenAI's Sora 2 Pro from a single screenshot of the costume. The creator told Reuters it was "intended as satire and demonstrates what AI video models can create with very little prompting or direction."
Target confirmed the ad was fake. The brand did the right thing and still spent the week answering for a commercial it never made.
Your crisis plan covers what you said. It doesn't cover what a stranger can generate in your name from one screenshot, by Tuesday.