1. Define Persona & Campaign Idea
Persona: Start by clearly describing your target audience (e.g., “Revenue Operations leaders in Europe”). Twain uses this to guide research and message generation.
Campaign Idea: Add a short description of what you want the campaign to focus on — e.g., interest in AI tools, simplicity in tool adoption, or pain points related to CRM complexity.
Best Practice: Keep your wording consistent. If you call your targets “leads,” always use “leads” (instead of mixing with “people” or “accounts”). Consistency ensures Twain generates messaging aligned with your terminology.
💡 A strong persona + campaign idea gives Twain the context it needs to generate relevant insights and personalized outreach.
2. Prepare Your Lead List
If you have a large list (e.g., 850 contacts), split it into two halves. This makes A/B testing easier.
Upload your leads into Twain.
Map essential identifiers such as:
Work Email (for delivery)
LinkedIn Profile (for research-based personalization)
Company Name, Title, First Name (optional, depending on how much data you want in the message).
💡 Clean, well-mapped data ensures Twain personalizes correctly for each lead and prevents errors during export.
3. Generate Messages
Once persona + leads are set, Twain can generate a first draft of your campaign messages.
Each draft contains:
Opener: Personalized hook or insight.
Body: Value proposition or research-based context.
CTA: What you want the lead to do (reply, book a call, etc.).
These are free to experiment with — credits are only used when you generate at scale for all leads.
💡 The first draft gives you a baseline. From there, you refine and improve to match your exact tone and goals.
4. Edit & Customize Messaging
Mix instructions with exact phrasing to shape Twain’s output. Example:
Instruction: “Make the tone conversational and short.”
Exact phrase: “Twain adds relevant deep research to outreach copy.”
Keep iterating until you’re happy with the sequence.
Adjust frameworks (e.g., pain-point > solution > proof, or research > question > CTA) to test different approaches.
💡 Editing ensures the AI’s draft sounds like you and communicates the right value. The framework flexibility lets you experiment with different narrative flows.
5. Set Up A/B Tests
Duplicate the campaign and change one key element (e.g., CTA or subject line). Examples:
CTA test: Campaign A asks for a call booking → Campaign B asks how they currently solve the problem.
Subject line test: Campaign A uses “Offer” subject lines → Campaign B uses “Source” subject lines (e.g., referencing an article the lead wrote).
Run the campaigns side by side.
Compare metrics such as:
Open rate (subject line effectiveness)
Reply rate (message resonance)
Positive reply rate (quality of responses).
💡 A/B testing shows which type of messaging actually works with your audience instead of guessing.
6. Export & Scale
Once you’ve finalized and tested messaging, generate messages at scale for all leads.
Export campaigns into:
Google Sheets → for easy review/editing.
Outbound Sequencers (e.g., Smartlead, Outreach, Apollo) → for direct sending.
Exports include:
Personalized subject lines (if enabled).
Full message sequences.
Custom variables (mapped fields).
💡 Exporting at scale allows you to launch campaigns while keeping all the personalization intact.
7. Founder’s Best Practices
Personas first: Invest time in creating detailed personas and campaign ideas. Twain’s personalization quality depends heavily on this.
Iterate quickly: Don’t aim for perfection upfront. Test drafts, learn from results, and refine.
Be opinionated: Use your own unique point of view in campaigns. The AI should sound like you, not just “any SDR.”
Stay flexible: Campaign performance data will guide you to the frameworks, CTAs, and subject lines that resonate most.