Skip to main content

​
Goal

Set up a playbook that follows up with contacts after a conversation outcome or a scheduled callback need.

​
Before you start

  • Decide which contacts or groups the playbook should target.
  • Decide whether the playbook should send messages, create reminders, or coordinate a later call.

​
Step 1: define the use case

Good first playbooks include:
  • callback scheduling after missed voice contact
  • post-demo or post-support follow-up
  • reminder outreach for unresolved cases

​
Step 2: create the playbook

In Playbooks:
  • name the playbook by business purpose
  • choose the cadence and timezone
  • keep the first version simple and observable

​
Step 3: connect the content

Use:
  • templates for email or SMS follow-up
  • conversation outcomes to choose who should enter the flow
  • contact groups when a campaign should target a known segment

​
Step 4: preview a launch batch

Publish your playbook, then open its launch form from Playbooks or Contact Groups.
  1. Select the source group and a contact limit, or choose All eligible contacts.
  2. Choose New contacts only (the default) or Include previous runs.
  3. Create a preview and review the selected contacts and exclusions.
  4. Click Launch reviewed batch to queue enrollment.
The limit counts eligible contacts. Selection starts with the oldest group membership and uses contact ID to break ties. A contact with an active run in any playbook is always excluded. New contacts only also excludes contacts who have run any published version of this playbook. Include previous runs permits contacts whose historical runs have ended. The preview freezes the cohort and published version. Later changes to group membership do not add or remove selected contacts. Changing the launch choices requires a refreshed preview. A preview with no eligible contacts cannot launch. Enrollment checks eligibility again. If a selected contact becomes ineligible, their result explains why they were skipped; the batch does not add a replacement.

​
Step 5: monitor enrollment and outreach

Open Batches from Playbooks, or follow the batch link from a run workspace. The batch workspace separates enrollment from outreach. Enrollment is complete when every selected contact has a result. Calls, conversations, reviews, and scheduled outreach can continue afterward. Use the contact table to inspect enrollment results, run status, current step, cycle, next scheduled action, and the latest outcome or error. Search by name, phone, or email, and filter enrollment and outreach independently. Results use 50-contact pages; summary totals always cover the whole cohort. Contact and run links open their existing workspaces. Historical batches show aggregate enrollment counts only. They do not claim ownership of old runs or offer batch controls.

​
Retry a failed step from human review

Open the pending decision in Human review or the run workspace. After resolving the cause of a failure, click Retry failed step to queue a new attempt of the action named in the decision panel. The original evidence and your notes remain in the history. Other pending reviews, pauses, and scheduling rules still apply. Retry is available when the failed action can run again without repeating later actions or approvals. It is unavailable for uncertain or partial delivery and expired scheduled cycles. You can still review the proposed outcome or use Manual override when appropriate.

​
Pause, resume, or cancel a batch

  • Pause batch blocks new enrollment and outbound work immediately. Linked runs update in the background, with progress shown on the page.
  • Resume batch removes only the batch pause. Individual operator pauses, stop requests, and pending reviews remain in effect.
  • Cancel batch permanently cancels pending enrollment and unfinished runs owned by this batch. Review the affected counts before confirming. Completed results remain available.
Calls already in progress and requests already submitted to a provider may finish. Delivery records remain available when a send was partial or its acknowledgement is uncertain. Batch control history records the operator and time.

​
Validate before increasing volume

First exercise the flow in your sandbox with mocked delivery. Then launch a small production cohort and reconcile the selected count against pending, enrolled, skipped, failed, and cancelled results. Check that each enrolled contact links to exactly one run owned by the batch. Inspect failures and pending reviews before increasing the contact limit. Batch tracking does not add click/open engagement triggers or payment processing.

​
Example workflow

  • a support call ends with “send docs and check back tomorrow”
  • the contact enters a playbook
  • the playbook sends follow-up content and schedules the next operational step
  • the team reviews run outcomes and adjusts the cadence if needed

​
What good looks like

  • the same task no longer depends on manual follow-up
  • run history is easy to audit
  • failures are obvious and actionable

​
Troubleshooting

  • If runs fail frequently, simplify the first version of the playbook.
  • If follow-up feels disconnected, align the playbook entry rules with conversation outcomes.