> ## Documentation Index
> Fetch the complete documentation index at: https://docs.recepta.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Campaigns

> Running a batch of contacts toward one goal.

A campaign takes a set of [contacts](/outbound/contacts), runs them through a [sequence](/outbound/sequences), and tracks what happened. It's the unit you launch, pause, and measure.

## Building one

<Steps>
  <Step title="Define the goal">
    One outcome per campaign — book a service visit, recover an unsold quote, sell a maintenance plan. A campaign with two goals converts on neither, because the message can't be sharp about either.
  </Step>

  <Step title="Select the audience">
    Filter contacts by tag, source, status, or custom field. Tighter is better: a hundred well-chosen contacts outperform a thousand vaguely relevant ones, and cost less.
  </Step>

  <Step title="Choose the agent">
    The agent that will make the calls. Its [knowledge](/receptionist/knowledge-base) applies as normal, so it can answer questions and book on the spot rather than promising a callback.
  </Step>

  <Step title="Attach a sequence">
    The steps that run per contact — calls, texts, emails, voicemail drops, and the waits between. See [Sequences](/outbound/sequences).
  </Step>

  <Step title="Set the schedule">
    When the campaign may contact people. Respect the contact's local time, not yours.
  </Step>

  <Step title="Test on a handful">
    Run ten contacts, ideally including your own number. Listen to a call and read a text before you release the rest.
  </Step>
</Steps>

<Warning>
  Calling hours are legally restricted — generally 8am to 9pm in the called party's local time, and tighter in some states. Set your schedule conservatively; a campaign that dials at 7:45am is a complaint waiting to happen. See [Recording & consent](/phone/recording-and-consent).
</Warning>

## Pacing

Don't launch a thousand calls at once. It floods your team with callbacks, concentrates any configuration mistake across your whole list, and looks like exactly what regulators watch for.

<Tip>
  Start at a volume your team can absorb if everyone calls back at once. Scale up when you've seen a full cycle complete cleanly.
</Tip>

## Monitoring

A campaign shows progress through its contact list, per-step outcomes, and the calls and conversations it generated. Watch the first hours closely:

| Signal                          | What it usually means                                       |
| ------------------------------- | ----------------------------------------------------------- |
| Very low answer rate            | Wrong time of day, or out-of-area caller ID                 |
| Answered but immediate hangups  | The opening isn't landing — read a transcript               |
| Good conversations, no bookings | The offer or the availability is the problem, not the agent |
| Complaints or opt-outs          | Stop the campaign and check your list's consent basis       |

<Note>
  Pause a campaign at any time. Contacts mid-sequence hold where they are and resume when you restart it.
</Note>

## Measuring

Judge on booked jobs and revenue. Contact rate and reply rate are diagnostics for fixing a campaign, not measures of whether it worked.

Calls generated by a campaign appear in your [call logs](/receptionist/call-logs) like any other, with full transcripts — the fastest way to understand why a campaign is or isn't converting.