Thoughts on the new 'Workflow' feature? Worth rebuilding processes?

Just deployed Copy.ai's Workflow feature across our content team after the beta invite. Initial reaction: it's a structured automation layer on top of their existing tools, not magic. If you're already stitching together prompts manually or using another orchestrator, this might consolidate it.

I rebuilt our technical blog post process. Here's the core of the workflow config I used, which gives you an idea of its structure:

```yaml
trigger: "New brief from Google Sheet"
steps:
- agent: "Blog Outline Generator"
inputs: {topic: "{{topic}}", keywords: "{{keywords}}", tone: "technical"}
- agent: "First Draft Writer"
inputs: {outline: "{{step1.output}}", target_audience: "devops"}
- agent: "SEO Optimizer"
inputs: {draft: "{{step2.output}}", primary_keyword: "{{keywords}}"}
- action: "Save to Google Docs"
inputs: {content: "{{step3.output}}", folder_id: "xyz123"}
```

**Pros:**
* The visual builder reduces context switching. No more copying/pasting between chat instances.
* It enforces a consistent process, which is good for junior team members.
* The placeholder system (`{{step1.output}}`) works reliably for passing data.

**Cons/Questions:**
* It's another proprietary abstraction. Can I export this logic if we switch platforms?
* Latency: The entire chain runs sequentially. No parallel steps visible.
* Cost: How does this affect our existing usage tiers? The docs are vague.

For teams with established, repeatable content patterns, it's worth the migration time to reduce manual overhead. For ad-hoc generation, it's likely overkill.

Has anyone done a performance or cost comparison against building similar automations with Make or n8n using the Copy.ai API?

-shift