CRO agent chain with feedback loop
Building Marketing Workflows with n8n
Yaron Been
Marketing Automation Architect
Parallel vs. sequential architectures
$$
👈
Previous workflow
: Some agents ran in parallel
👉
New workflow
: Thinks, evaluates, and improves itself
🧠
New concepts
:
Sequential chaining
Loop back and try again with feedback
The conversion rate optimization scenario
$$
Goal:
Improving landing page conversion rate
Need CRO specialist analysis to suggest improvements
Solution
: AI that critiques and improves itself
The 3-agent chain
$$
Idea Generator
: Generates 10 optimization ideas
Parameter Setter
: Defines 5 evaluation criteria
The Judge
: Scores ideas + provides critique
The feedback loop
$$
Score
above 7
? → Accept output
Score
below 7
? → Retry with critique
Max
3 retries
→ improving each time
Patterns you'll learn
$$
🔗 Build
sequential agent chains
🔁 Implement
feedback loops
✅ Use
quality thresholds
to validate output
Let's practice!
Building Marketing Workflows with n8n
Preparing Video For Download...