Framework

Decision Stack Canvas

Map what your agent-first purchase flow needs to do — and what your product must provide — across the five stages.

How to Use

The Five Stages

1. Intent

What need triggers the purchase? What constraints matter? (budget, timing, preferences)
Notes / answers:
Signals the agent can infer
Signals you should collect explicitly

2. Research

Where does the agent look? What structured data should it find about you?
Notes / answers:
Databases/aggregators to be present in
Schema/feed fields required

3. Evaluation

How does the agent compare options? What differentiators matter for your best-fit segment?
Notes / answers:
Proof and reputation signals needed
Key tradeoffs to win

4. Transaction

How does an agent buy from you? What makes checkout safe and low-friction?
Notes / answers:
APIs/auth standards supported
Guardrails / limits / approvals

5. Fulfillment

What happens after purchase? How do you reduce returns and increase satisfaction?
Notes / answers:
Delivery and tracking signals
Return process and support

Want Feedback on Your Canvas?

This canvas is from Instant Checkout: How AI Agents Are Quietly Replacing the Way We Buy