Wizard Step 1: Basics

This step decides your campaign identity: what it’s called and which agent runs every call. If you get this wrong, everything downstream becomes harder to debug.

Campaign wizard Basics step with template, campaign name, and agent selection.

At a glance

  • You can optionally prefill defaults from a template.
  • You must set Campaign Name and Agent (required).
  • The agent defines the script, tools, and voice used on every call.

Before you start

  1. Know the goal in one sentence (example: “Book a demo for qualified inbound leads who went cold”).
  2. Know the audience segment (who should receive this call).
  3. Pick one primary agent that matches that audience and goal.
Pro tip: name campaigns like you’ll debug them later

A good name makes reporting and handoffs easy. Use a consistent pattern like: [Audience] - [Offer] - [Geo] - [Month].

Field-by-field (what it means + how to decide)

Load from Template

Recommended Use templates when your team wants a standard operating baseline (same schedule, concurrency, retry settings).

Use templates when:

  1. You run similar campaigns repeatedly.
  2. You want consistent defaults for new operators.
  3. You maintain “gold standard” templates quarterly.

Don’t use templates when:

  1. You’re cloning a proven campaign (use Clone instead, so you copy the real-world configuration).
  2. The audience or goal is new and needs bespoke settings.

Campaign Name (required)

Recommended Make the name understandable without opening the settings.

Good examples:

  • Inbound Leads - Demo Follow-up - US - Apr 2026
  • Reactivation - Past Trials - UK - Week 17

Bad examples:

  • Campaign 12
  • Test

Agent (required)

Recommended Pick the agent that matches the audience + intent + follow-up path.

This choice affects:

  • what the AI says (script and opening),
  • what the AI can do (tools/actions),
  • how outcomes are interpreted.
Watch out: “one agent for everything” creates generic calls

If one agent handles multiple audiences, the opening becomes vague and objection handling becomes inconsistent. Split agents when the first 15 seconds of the call should sound meaningfully different.

Validation rules (what blocks Next)

You can’t proceed unless:

  1. Campaign Name is not empty.
  2. Agent is selected.

Common mistakes (and quick fixes)

  1. The agent sounds wrong for the audience. Fix: choose a more specific agent, or split into separate agents by segment.
  2. Naming is inconsistent. Fix: adopt one naming standard and apply it across all campaigns before you scale.
  3. Template drift. Fix: treat templates as owned assets (quarterly review, one owner, published version history).

Final checklist (what good looks like)

  • The name explains audience + offer + geography + time window.
  • Everyone agrees the agent fits this campaign goal.
  • You can answer: “What should happen after a good call?” in one sentence.
Next: choose your numbers
Go to Telephony and build a number pool that can deliver reliably.

Additional resources

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.