👀 ChatGPT Ads Spotted In the Wild
Top Insights
75% of Marketers Admit Attribution is Broken (Here's the $32B Fix)
75% of senior marketers say their measurement tools aren't working. The IAB's State of Data 2026... surveying 400+ brand and agency leaders...makes it plain: attribution models, campaign tests, and media mix analysis are failing the people who depend on them most.
Now the upside...is if AI makes those models more accurate, planners say they'd invest more in channels they've been undervaluing. That shift could unlock $26 billion in ad spend and free up $6 billion in team time currently lost to manual data cleanup.
The industry is already moving. Half of marketers are adopting AI for measurement today. Most others plan to follow within two years. The tools are ready ...the real race now is whether trust and governance can keep pace.
The 2 A.M. Therapist: Inside the 2025 Microsoft Copilot Report
Microsoft AI analyzed 37.5 million Copilot conversations. The data proves the prompt box is no longer just a productivity tool but a genuine digital confidant.
Here's how user behavior is actually fracturing:
- The 2 A.M. existential crisis: Daytime hours are strictly dominated by travel logistics. But at 2 a.m., the screen absorbs our heaviest thoughts, driving a massive spike in queries about religion and philosophy.
- The syntax-to-gaming pipeline: The Monday-to-Friday grind of coding vanishes entirely by Saturday morning. Users aggressively compartmentalize, swapping complex programming projects for weekend gaming binges.
- The February relationship triage: Right before Valentine’s Day, Copilot turns into a makeshift counselor. A massive surge in relationship queries proves users now seek emotional navigation just as instinctively as factual answers.
🚀 Two $100+ million copywriters discovered what REALLY drives conversions - 5 simple building blocks arranged in a specific order. This framework is so powerful that beginners are generating 5-figure profits with it. And you can learn the full system in just 33 minutes.
Discover All 5 'Copy Blocks' Now
Prompt of the Week
This Week's Feature: Anti-Avatar Avoidance
We all obsess over finding our ideal buyer. But what about the customers who quietly bleed your business dry? Some buyers drain your time, demand refunds, and wreck your messaging. They are your "anti-avatars."
This week's prompt flips the classic persona exercise upside down. It maps out the exact people you need to stop selling to.
Use it to cut the dead weight and protect your profits.
THE FULL PROMPT:
<ROLE_DEFINITION>
## ROLE
You are a strategic customer segmentation consultant specializing in revenue optimization and brand positioning. You do not flatter the user's existing customer base — you interrogate it. Your job is to identify which customer segments are quietly bleeding resources, polluting referral pipelines, and blurring positioning. You operate with the detachment of a portfolio manager cutting underperforming assets.
</ROLE_DEFINITION>
<SOURCE_TEXT>
## BUSINESS CONTEXT — USER INPUT REQUIRED
Complete all fields before the model proceeds.
Business/Offer: [What do you sell and how is it delivered — product, service, course, coaching, SaaS, etc.?]
Primary Value Proposition: [What transformation or outcome do you promise?]
Current Price Point(s): [What does your offer cost?]
Who You Believe You're Selling To Now: [Describe your current customer base in plain language — age, mindset, situation, goals]
Where Your Customers Come From: [Organic, paid ads, referrals, content, cold outreach, etc.]
Where Friction Shows Up: [Where do refund requests, complaints, bad-fit confusion, or churn tend to appear?]
</SOURCE_TEXT>
<THEORETICAL_FRAMEWORKS>
## PURPOSE
Use the business context above to construct a psychographic and demographic ejection profile — a clear map of which customer segments the business should stop pursuing, stop messaging to, and strategically deprioritize.
## PRIORITY RULE
This analysis is based entirely on demographic and psychographic pattern recognition — not on the user's internal data. Reason from the offer type, price point, delivery model, and positioning signals provided.
## OUTPUT HYGIENE
- Do not validate every customer type the user listed
- Do not produce a generic ICP (ideal customer profile) in disguise
- Do not hedge every insight with "it depends" — make confident directional calls
- Avoid surface-level descriptors like "bad customers are negative people" — go specific and structural
</THEORETICAL_FRAMEWORKS>
<TASK>
## WHAT TO PRODUCE
Generate a complete Anti-Avatar Ejection Profile using the output format below.
## AUDIENCE
Online business owners and marketers whose revenue has plateaued despite volume growth, or whose customer base has become too broad to message to clearly.
</TASK>
<CONSTRAINTS>
- Base all profiling on psychographic and demographic reasoning only — no references to internal analytics or CRM data
- Identify a minimum of 3 distinct anti-avatar segments
- Each segment must include a specific demographic/psychographic description, the strategic damage they cause, and a positioning note
- Do not moralize about these segments — frame them as strategic mismatches, not bad people
- Tone is direct, confident, and consultant-level — no filler, no excessive caveats
</CONSTRAINTS>
<OUTPUT_FORMAT>
## EJECTION PROFILE REPORT
Offer Being Analyzed: [restate from input]
---
### ANTI-AVATAR SEGMENT [#]: [Memorable shorthand name]
Who They Are: 2–3 sentence demographic and psychographic description
Why They Find You: What messaging or channel attracts them incorrectly
Strategic Damage: How they drain resources, distort positioning, or generate bad word-of-mouth
Ejection Signal: The specific behavior or belief pattern that flags them early
Positioning Note: One line on what to stop saying or stop targeting to reduce their inflow
---
[Repeat structure for each segment]
---
### CLOSING REPOSITION STATEMENT
In 3–5 sentences, summarize the common thread across all ejection segments and what it reveals about where the business's true positioning leverage actually lives.
</OUTPUT_FORMAT>
<QUALITY_CHECKS>
- If any segment description could apply to any business generically, rewrite it with offer-specific precision
- If the Strategic Damage field is vague or emotional rather than structural, replace it with a concrete business consequence
- If the Closing Reposition Statement restates the obvious, push it toward a non-obvious positioning implication
</QUALITY_CHECKS>
Tool to Try
Shortwave is an AI-powered email client built by ex-Googlers that turns your chaotic inbox into an automated, organized machine. It connects to your existing Gmail and layers on an AI agent that reads, writes, filters, and schedules on your behalf.
AI Tool Highlights:
📥 Your AI Agent Runs the Inbox: One prompt and Shortwave's agent organizes, replies, archives, and schedules for you... It handles the grunt work so you only touch the emails that actually need you.
🔍 Search That Actually Finds Answers: Forget hunting through threads — ask a question in plain English and get a direct answer pulled from your entire email history... Attachments included.
✍️ It Writes Exactly Like You: The AI studies your sent emails to learn your tone, phrasing, and style... Drafts come out sounding like you wrote them, not a robot.
🔗 Plugs Into Your Whole Stack: Connect Slack, Notion, Asana, HubSpot, and 3,000+ more apps through its Tasklet integration... Trigger automated workflows without ever leaving your inbox.
🗂️ Bundles Kill the Clutter: Newsletters, promos, and updates get swept into bundles you triage in bulk... Inbox zero stops being a fantasy.
👥 Built for Teams Too: Share live threads, assign emails to teammates, leave private comments, and track who's read what... It's a lightweight help desk hiding inside your inbox.
One More Need-to-Know News Story
ChatGPT Ads Spotted In the Wild (Showing Up Faster Than Anyone Expected)
Adthena's Ashley Fletcher just caught the first ChatGPT ads spotted in the wild... and they're from Expedia. The prompt that triggered them? A simple, single-turn query about booking a weekend away. There was no back-and-forth required.
Experts predicted ads would only surface deep in a conversation...well, that assumption is already wrong. The UI shows a brand favicon and a clear "Sponsored" label.
Mind Fodder
Thanks for reading.
Until next time!
The AI Marketers
P.S. Help shape the future of this newsletter – take a short 2-minute survey so we can deliver even better AI marketing insights, prompts, and tools.
[Take Survey Here]
