Visa wants to give AI Agents "tokens" so they can pay without you ever seeing a checkout page.
Visa's CEO told investors this is their #1 priority.
Here's how it will work 👇
Imagine telling your Agent.
"I want a signed Pedro Pascal shirt from The Last of Us" don't spend more than $100" -
Then you tap your phone.
The next thing you know, the payment is complete.
Let's unpack the flow as I understood it 👀
1. Tap a card to a device to get a token for their AI Agent 2. The AI Agent registers with the network as a "trusted wallet" in the same way Apple Pay can register with networks and processors 3. The user gives the AI Agent a mandate (Find a Pedro Pascal shirt, don't spend more than $100) 4. The Auth instructions are finalized ($100 budget, Pedro shirt) 5. The Auth instructions define when step-up or biometric authentication is required (e.g. > $100, or some level of fraud risk detected) 6. The auth instructions are matched between the Agent, and the network at the Merchant (e.g. This agent has a mandate to buy for a max of $100 this SKU) 7. The payment is complete. BUT, Liability remains with the merchant for fraud (!!)
You as a customer you could give your AI agent any kind of mandate.
Like,
- Here's $2,000 for monthly groceries.
- Here's $20,000 to book and run ads this month
- Here's $100,000 to hire AI Agents for engineering
Wild. But more so for merchants.
The implications for merchants and acquirer processors are insane.
- Every A/B test you're running? Obsolete.
- Cart abandonment emails? Pointless.
- Conversion optimization? For agents, not humans.
Bowing to pressure, Goldman Sachs will no longer pursue checking accounts for the mass market. Bloomberg reports Goldman will also limit their consumer lending but will continue their existing card partnerships. 🧵
🤔 There goes the case study. When pointing at "who's doing something interesting in consumer, every consultant (including me) would point at Marcus.
Amazing resilience and case study in how to do "new" in big org. Reality bites. Massive org w/ shareholders and P&Ls has bitten.
🤔 It was a great strategy but expensive execution. Marcus had three legs to their strategy. 1. Becoming a universal bank direct to the consumer across all product types. 2. Partnerships like Apple Card. 3. Starting to offer APIs to get into Banking-as-a-Service.
The average P/E multiple for Fintech businesses in public markets is 3.4x. That puts the Fast .com's 166x revenue multiple from 2021 into a sharp perspective!
So is Fintech dead? I don't think so.
The whole venture and tech sector is down
Of the 124 tech IPOs in 2021, only 15 are above their trading price
Fintech deal velocity is also slowing (h/t @ftpartners)
*Some* Fintech companies burned cash rapidly to compete on who could grow fastest.
With the pandemic and unlimited marketing spend, many hired too quickly, and did so with questionable future business models.