Learn and master Clay with Clay university

In the last lesson we covered AI Snippets. These are a modular, high-control approach to generating content where you can customize every part of a message. But what if you don't want or need to customize every step?

What if you're looking for a lower-friction way to start your outbound sequences… and just want the AI to generate full email copy for you?

That's exactly what the Twain enrichment is for.

🧩 How Twain Structures Your Messaging

Instead of asking AI to generate a single block of text, Twain walks the model through a multi-step process, typically including:

• Subject Line
• Intro Line
• Body or Value Prop
• Call to Action (CTA)
• PS or Signature (optional)

And it can automatically use the enriched data in your Clay table, like:

• Company name and description
• Job title
• LinkedIn bio
• Recent funding
• Tech stack
• Or any other signals you've pulled in

All that data gets plugged into a structured template behind the scenes.

✅ When to Use Twain

Twain is a good solution when:

• You want full email copy, not just snippets
• You need to ensure you're not overloading AI with too much unstructured data
• You still need to customize tone, length, or structure
• You want outputs that are clean, modular, and easy to edit

🔄 Using Twain with Your Research Data

Let's go back to our previous example.

We've got our leads and we've done some research. Now we want to start sending personalized outbound messages.

Using the same leads we want to contact, let's look at what this looks like with Twain.

🛠️ Step-by-Step Setup Process

Here's how to set up your sequence:

1. Choose your sequence type:
   • Select a pre-defined sequence (1-4 steps) from the dropdown menu
   • If you have a Twain account, you can use your own sequence by selecting your Twain account and entering a sequence ID
2. Configure your prompt parameters:
   • Add your prospect description
   • Define your outreach signal
   • Describe your ideal customer profile
   • State the problem you're solving
   • And more…
3. Add your prospect information:
   • Enter first name, last name, and LinkedIn profile URL for each prospect
   • Don't have LinkedIn URLs? Use Claygent to find them using prospect emails
4. Map and execute:
   • Map Twain's multi-step messages to your Clay columns
   • Run the table to generate your sequences

🤔 AI Snippets vs. Twain: When to Use Each

This might leave you wondering – when should you use AI snippets vs Twain?

• Use AI Snippets when you want full control and flexibility at the sentence level
• Use Twain when you want a faster, more structured way to generate outbound messaging at scale

Both are powerful. Choose the one that best fits your workflow. Experiment with both. Scale what works.

💼 Benefits of a Multi-Step Approach

The multi-step approach that Twain uses offers several key advantages:

1. Better structure: By breaking the generation process into discrete steps, the AI focuses on one task at a time, producing more coherent results
2. Improved personalization: Each step can incorporate different aspects of your prospect data
3. Easier editing: If you don't like a particular part of the message, you can regenerate just that section
4. Sequence consistency: Ensures your follow-up messages maintain a consistent voice and approach

🌟 Real-World Applications

Beyond basic outbound sequences, you can use Twain for:

• Multi-touch campaigns that gradually build on previous messages
• Post-event follow-up sequences
• Re-engagement campaigns for dormant prospects
• Content distribution sequences that share different resources over time

🔮 What's Coming Next

We've covered effortless output when it comes to writing copy. There's still a number of creation workflows, like video, slides, and more. We'll cover those in future lessons.

As you move forward with your AI-powered GTM automation, remember that the tools and techniques we've covered in the FETC framework work together as a system. The Find, Enrich, and Transform steps all feed into your Create step, giving you the data and structure you need to generate high-quality, personalized content at scale.