In partnership with
Happy Monday.
Nobody in marketing owns the wallet. It is the last screen, somebody in payments chose it years ago, and it reaches your calendar only when conversion dips and everyone wants to know whose fault that was.
On the 18th Stripe wired its Link wallet into Meta's Muse, xAI's Grok and a texting assistant called Instinct, so an agent can pay on your customer's behalf. Link has 300 million users and sits on around a million merchant checkouts already, Payments Dive reported. It mints a single-use card for each purchase, so the bot never touches the real number.
I know I led on Mastercard giving bots a virtual card eight days ago, and I nearly skipped this one for that reason. Then it landed. A card network issuing plastic to a robot is a pilot you can wait out. A wallet already living inside your checkout quietly accepting a robot is a default. Nobody schedules a procurement meeting about a default.
Richard Crone, who runs a payments advisory firm and discloses PayPal as a client, put the shape of it in one line: "Essentially, the way to think about this is that the agent will be the wallet or the wallet will be the agent."
Which is how these rails always get laid. By people with no opinion whatsoever about your funnel.
Nobody is costing what it does to the work. A capped single-use card does not feel urgency, does not see your retargeting, and does not abandon a basket you can rescue with a discount email.
It was never hesitating. It was executing, and most of your paid budget is built to argue with somebody having second thoughts.
The session you would use to prove any of this never happens either. No landing page, no scroll depth, no form fill. The order lands in revenue with nothing attached to it.
So, your Monday question. If a slice of next year's volume arrives from something that read your product page once and never came back, what in your reporting would even flag it?
Today’s sponsor:
Some teams never seem to stop moving. They're on Attio, the agentic CRM.
Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.
With Attio, you’ll get:
Leads automatically prioritised and routed to the right rep
Expansion and risk signals caught the moment they land
Follow-ups written in your voice, already there when you arrive
Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?
TALK ON THE MARKET
Joe Futty, chief product and technology officer at Pipedrive, launching its meeting intelligence tool, in Solutions Review's weekly round-up:
Nova gets everything else out of their way so reps can spend that energy doing the one thing they're actually paid to do: sell. That's what we built Nova to do.Alyson Griffin, Head of Marketing at State Farm
Every vendor is selling that sentence this quarter. We get things out of your way.
Stripe just got the buyer out of your way, and nobody asked marketing first.
Want to see case studies of AI-powered campaigns?
NUMBERS MATTER
46%. The share of marketers who check whether AI answers mention their company by hand, prompt by prompt, in a live poll run at HubSpot UNBOUND. Another 43% have a tool for it.
Print the sample size with it: 35 people, in a room, polled by the company that sells earned-media distribution. Small, and interested in the answer.
It still names the problem. Most of the industry is monitoring its newest channel the way you would check a rumor.
DAILY PROMPT
Our ICP Refinement Through AI-Powered Market Research, for the planning cycle where your buyer might not be a person:
You are refining an Ideal Customer Profile (ICP). Work through the inputs below, read any files attached to this conversation (CRM exports, win/loss notes, call transcripts, pipeline reports), and research the current market on the web where the data is thin. State which conclusions come from my data and which come from outside research.
## Current ICP Definition
Provide your existing ICP:
- Target industry/vertical: [INDUSTRY]
- Company size (employees/revenue): [SIZE_RANGE]
- Job titles/personas: [KEY_PERSONAS]
- Geographic focus: [REGIONS]
- Current pain points addressed: [PAIN_POINTS]
## Research Inputs
Analyze the following data sources to refine the ICP:
### Customer Data
- List 3-5 of your best/most profitable customers with brief context: [BEST_CUSTOMERS]
- What do they have in common beyond the obvious? [COMMONALITIES]
- Customer acquisition cost and lifetime value for top segment: [CAC_LTV]
### Market Intelligence
- Recent industry trends affecting your target market: [MARKET_TRENDS]
- Emerging job titles or organizational structures in your space: [ORG_CHANGES]
- Competitive landscape shifts: [COMPETITIVE_SHIFTS]
### Sales/Support Feedback
- Most common objections from non-ideal prospects: [OBJECTIONS]
- Sales team observations about who buys vs. who doesn't: [SALES_INSIGHTS]
## Analysis Framework
If I attach raw account data, segment it yourself and run the math rather than relying on my summary. Then provide:
1. **ICP Refinements**: Identify 2-3 specific ways to narrow, expand, or redefine the ICP. For each refinement, explain the business rationale and the evidence behind it.
2. **Emerging Segments**: Identify any secondary ICPs or high-potential segments that differ from the primary ICP.
3. **Disqualification Criteria**: Define 3-4 clear characteristics that indicate a prospect is NOT a good fit.
4. **Go-to-Market Implications**: How should messaging, positioning, or channel strategy shift based on ICP refinements?
5. **Validation Recommendations**: Suggest 2-3 ways to test these ICP refinements with real market data before major GTM changes.
Flag where the sample size is too small to support a conclusion. Structure your response with clear headers and bullet points for easy scanning and team sharing.Run it before the 2027 planning meeting, then add the question it does not ask. For each segment, could an agent be standing in at the moment of purchase, and what would it need to find on your page to say yes?
CMO CORNER
If your team relaxes when you're away, something is broken.
The same test works on a stack. A checkout that behaves one way while you watch and another at 3am is unattended rather than automated, and those two words carry different insurance.
Ask for one number this week. What share of completed orders had no human session attached. If nobody can produce it, you have your answer.
We provide consulting services for CMOs on AI strategy for marketing teams through our Teamless offering.
TOOL UPGRADE
If the buyer is arriving as software, three places to put software of your own:
Chatbase — best for standing up a branded assistant that answers product questions in your words, before an agent goes and guesses them. Try Chatbase here →
Lemlist — best for outbound teams who want sequences personalized per account without hiring another SDR. Try Lemlist here →
Reply — best for running that motion across email, LinkedIn and calls once the list outgrows the team. Try Reply 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
Before you buy anything that touches the checkout: AI Integration Planning Framework for Martech. How to evaluate, sequence and actually land AI across a stack you already own.
Read the prioritization section before the next vendor call. The pitch will be about capability, and your problem is order of operations.
→ We publish workshops, courses, guides, manuals, and more on AI marketing for free. Browse our learning resources for free.