Before You Start
Make sure the following are in place before creating a campaign:
At least one identity connected with LinkedIn and/or email (see the Enginy Configuration Guide).
Email account connected if your sequence includes email actions. Without it, email steps will block the progress of leads.
Contact list ready with enriched contacts (see the List Building Guide).
SPF, DKIM, and DMARC configured on your sending domain (for email campaigns).
Email warm-up completed for new email addresses or domains (see the Enginy Configuration Guide).
Prepare AI Variables for Message Generation
If you want to generate unique, personalized messages with AI for every prospect, you need to create AI Variables before building your campaign.
What Are AI Variables
An AI Variable is a reusable prompt that tells Enginy how to research and return a specific piece of information for a Contact or Company. Think of it as a smart column: the AI reads the contact's attributes (name, job title, company, LinkedIn activity, etc.), optionally searches external sources, and returns the answer in a format you specify.
AI Variables are used for two purposes in campaigns:
Message generation — Create an AI Variable for each touchpoint in your sequence. For example, a 5-step campaign (Email, Email, LinkedIn, LinkedIn, Email) would need 5 AI Variables, one per message.
Data enrichment — Generate additional data points for segmentation or qualification (covered in the List Building Guide).
Create an AI Variable
Navigate to AI Playbook > AI Variables.
Click Create AI Variable.
Select the entity type: Contact or Company.
Configure the variable:
Field | Description |
Title | Name of the AI Variable. Cannot be changed after creation. |
Output type | Format of the result: Text (messages, summaries), Number, Date, One of (predefined tags like Yes/No or Hot/Warm/Cold), URL, or Email. |
Prompt | The instruction that defines persona, context, goal, and formatting rules. Use |
Deep Research (optional) | Enables the AI to browse the web and extract data not available in your lists. Increases cost. |
AI Model | Choose the model (e.g., GPT 5.2, Grok 4, Gemini 3 Pro, Claude 4.5 Sonnet). Models marked Recommended are best for general use. |
Recommended prompt structure for messages:
Persona — Define the role the AI represents.
Context — Include sender variables (your company info) and receiver variables (contact data).
Goal — What the AI should produce.
Instructions — Greeting format, icebreaker options, value proposition, call to action, tone, and length constraints.
Signature — Specify that the sign-off should be omitted (for emails, the signature is added automatically from the identity configuration).
Examples — Provide 1–2 examples of the desired output.
Tip: Use the Enhance button below the prompt editor to let the AI polish your prompt automatically. You can also start from pre-built Templates for common use cases.
Note: For the full prompt best practices guide, available AI models comparison, Deep Research capabilities and limitations, and advanced use cases, see the AI Playbook article.
Run AI Variables on Your List
Once created, AI Variables appear as columns in your lists. Run them before launching a campaign so that each contact has its personalized message ready:
Open the list and locate the AI Variable column.
Click the column header and select Run column > All rows (or select specific rows first).
Alternatively, click Enrich > Enrich with AI and select the variables to run.
Important: AI-generated messages in campaigns do not have an extra credit cost beyond the AI Variable generation itself.