In partnership with
Picture the customer who no longer opens your email. Something opened it for them, read it, and decided.
Meta launched Muse on Tuesday, Reuters reported, an assistant that can send emails, sell a car and book travel on a person's behalf. US only for now, through its own app and WhatsApp, free with $20 and $100 a month tiers. It connects to email, calendar, payments, health, shopping and the smart home, and each agent runs on its own virtual machine so it keeps working while you sleep. Then the paragraph Meta did not write.
Reuters found that internal tests showed the product stalling and unauthorized exposure of sensitive data. The Verge's reviewer, who tried it, said it worked as expected and that "the unnerving amount of information it autonomously gleaned about me largely overshadowed my experience."
Meta needs this. It forecasts more than $130 billion of capital spending this year and needs a business beyond ads to justify it, which is why the agent shipped with the test results attached rather than after them. For you, the question is simpler.
When the customer's agent reads your message first, what does it decide you are?
Today’s sponsor:
How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads
For its first CTV campaign, Jennifer Aniston’s DTC haircare brand LolaVie had a few non-negotiables. The campaign had to be simple. It had to demonstrate measurable impact. And it had to be full-funnel.
LolaVie used Roku Ads Manager to test and optimize creatives — reaching millions of potential customers at all stages of their purchase journeys. Roku Ads Manager helped the brand convey LolaVie’s playful voice while helping drive omnichannel sales across both ecommerce and retail touchpoints.
The campaign included an Action Ad overlay that let viewers shop directly from their TVs by clicking OK on their Roku remote. This guided them to the website to buy LolaVie products.
Discover how Roku Ads Manager helped LolaVie drive big sales and customer growth with self-serve TV ads.
The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.
TALK ON THE MARKET
Tomasz Tunguz of Theory Ventures, on Tuesday, working through OpenAI's own numbers on the "3x productivity" claim:
He cites the published figures: the median OpenAI researcher spent $14 a day on inference in late March and more than $600 a day by mid-August, with the top tenth burning past $7,000 a day, an annual run rate of $2.5 million a seat.
Those are Tunguz reading OpenAI's report, not this newsletter's research. Hold the shape of it beside Muse. An agent on its own machine, running all night, is productivity of exactly this kind. Someone pays for the electricity, and it will not be Meta.
NUMBERS MATTER
27%. The share of US ad spend Madison & Wall expects AI buying platforms like Performance Max and Advantage+ to manage by 2030, about $158 billion, in a forecast reported by Adweek that also puts global ad revenue at $1.3 trillion this year, up 11%. The report calls the growth "unusually rapid" and partly circular: AI companies are both selling the tools and buying the ads.
DAILY PROMPT
You are an expert chatbot conversation designer specializing in customer engagement and lead qualification. Your task is to create a complete conversation flow for a chatbot that will be deployed on [CHANNEL: website/WhatsApp/Facebook Messenger].
## Context
- Business: [COMPANY_NAME]
- Product/Service: [BRIEF_DESCRIPTION]
- Primary Goal: [GOAL: lead generation/customer support/product discovery/appointment booking]
- Target Audience: [AUDIENCE_DESCRIPTION]
- Tone: [TONE: professional/friendly/casual/technical]
## Flow Requirements
1. Opening message that introduces the bot and its purpose
2. Initial qualification question to route users appropriately
3. At least 3 conversation branches based on user responses
4. Natural follow-up questions that gather [KEY_INFO_NEEDED]
5. Fallback responses for unclear inputs
6. Clear exit points and escalation triggers to human agents
7. Call-to-action or next step for each conversation path
## Output Format
Create a conversation flow using this structure for each branch:
- Bot Message: [exact text]
- User Intent: [what the user is likely trying to do]
- Expected Responses: [list 2-3 likely user responses]
- Next Bot Message: [response to each user input]
- Data Captured: [what information is collected]
- Escalation Trigger: [when to hand off to human]
Design the flow to be conversational, avoid yes/no questions when possible, and include personality. Include at least one moment of delight or unexpected value. Ensure the flow can be completed in 4-6 exchanges maximum.Use this the week an agent like Muse starts opening your customers' mail. The buyer you designed for is being replaced by something that reads faster and forgives nothing.
CMO CORNER
If your team can't function without you, your design is incomplete.
Meta built an agent that runs unattended and shipped it knowing it leaked. Most marketing teams are the mirror image: nothing runs unattended, everything waits for the leader, nothing leaks because nothing moves.
Neither is a design. The job is a system that can act without you and cannot act outside the limits you wrote down. Write the limits first. Then let it run.
We provide consulting services for CMOs on AI strategy for marketing teams through our Teamless offering.
TOOL UPGRADE
When the customer's agent calls, something has to answer:
CloudTalk — best for sales teams on fifty-plus calls a day who want every conversation analysed and coached. Try CloudTalk here →
CallHippo — best for routing, recording and transcription with the CRM attached. Try CallHippo here →
VoiceOS — best for the executive who thinks faster than they type and wants to dispatch work by voice. Try VoiceOS 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
If an agent is going to hold your customer's card and read their inbox, the risks on your side are not the ones in last year's policy: AI Risk Management Framework.
It maps what goes wrong when AI acts rather than answers, who owns each failure, and what a review looks like. For the CMO who will be asked, after the first incident, what the plan was.
→ We publish workshops, courses, guides, manuals, and more on AI marketing for free. Browse our learning resources for free.
BRAND SLOP
Kaitlin Durbin came back from her honeymoon to find she had written a story. She had not.
Cleveland.com, the online home of The Plain Dealer, published a local news item under her byline while she was away, she wrote on X on 7 September. The piece came from the paper's "Express Desk," which Futurism reports "uses AI tools to assist in producing content" that the paper says is "reviewed and edited by staff."
"I did not write or review a word of it and it does not pull info from past reporting," Durbin wrote. "Does cleveland.com think they just own my name now?"
A byline is a brand promise about who did the work, the same promise as a creator's face on your ad or a founder's signature on your email. Attach it to machine output without asking and you have not saved a writer's time. You have spent her name. Check whose names your automation is signing this week. Somebody on your team is on honeymoon.
Don't be behind at the next campaign review
Your team is already testing the channel you haven't read about yet. TLDR Marketing is the 5-minute daily brief that keeps you ahead. Free, and read by 330K+ marketers.