Creating a Custom Playbook

Last updated: January 24, 2026

Overview

Custom playbooks allow you to build tailored accounts receivable automation workflows from scratch. Unlike pre-built templates, custom playbooks give you complete control over the conditions and actions that determine how your collection conversations are handled.


Step 1: Create a New Playbook

Navigate to Playbooks from the main navigation menu, then click the Create Playbook button in the top right corner.

Create Playbook Form

Fill in the playbook details:

Field

Description

Playbook Name

A descriptive name for your workflow (e.g., "High Value Escalation Workflow")

Description

Brief explanation of what the playbook does

Customers

Choose "All customers" or select specific customers

Click Create Playbook to proceed to the workflow editor.

Important: Each customer can only be assigned to one active playbook at a time. If a customer is already in another playbook, they will appear as "In other playbook" and cannot be selected.


Step 2: Access the Workflow Editor

After creating your playbook, you'll be taken to the Workflow Editor. This is where you define the logic that determines what happens during collection conversations.

The editor has three main areas:

Area

Purpose

Main Canvas

Visual representation of your workflow steps

View Toggle

Switch between "Nested" (tree view) and "Flow View"

Configuration Sidebar

Edit details of selected steps


Step 3: Add Your First Step

A custom playbook starts empty. Click Add Step to add your first step:

Option A: Add a Condition

Conditions evaluate data about the invoice or conversation and branch into TRUE/FALSE paths.

  1. Click Add Step > Add Condition

  2. The condition appears in the canvas

  3. Click the condition to configure it in the sidebar

Option B: Add an Action

Actions perform operations like sending emails or escalating conversations.

  1. Click Add Step > Add Action

  2. Select the action type

  3. Configure action settings in the sidebar

Best Practice: Most playbooks start with a condition to evaluate the situation, then branch to appropriate actions.


Step 4: Configure AI Conditions

AI conditions use natural language to evaluate conversation content and customer sentiment.

Condition Type: AI

  1. Select the condition step

  2. Set Condition Type to "AI (Natural Language)"

  3. Enter a descriptive prompt in natural language

Writing Effective AI Prompts

The AI has context on the conversation history, invoice details, and customer information. Write prompts that clearly describe when the condition should evaluate to TRUE.

Good examples:

  • "Set to true if the customer mentions wanting a payment plan, payment arrangement, or requests to split the payment"

  • "Set to true if the customer claims they have already paid this invoice"

  • "Set to true if the customer disputes the charges or questions the invoice amount"

Avoid vague prompts:

  • "Check if customer is unhappy" (too vague)

  • "Payment issues" (not a complete condition)


Step 5: Add Actions to Condition Branches

Each condition has TRUE and FALSE branches. Add actions or nested conditions to each branch.

Adding Actions to a Branch

  1. Find the branch label (TRUE or FALSE)

  2. Click Add Step next to the branch

  3. Select an action type

Available Actions

Action

Description

Configuration

Escalate Conversation

Flags for manual review

Optional reason text

Send Collection Email

Sends automated collection email

Tone (friendly/professional/firm), Additional recipients

Loop in Email

Adds CC recipients to future emails

Email addresses


Step 6: Test Your Playbook

Before activating, test your playbook to verify it works as expected.

  1. Save your changes by clicking Save

  2. Click the Test dropdown

  3. Select a specific customer or let Monk choose randomly

  4. Review the test results

Understanding Test Results

The test runs in dry-run mode - no actual emails are sent or conversations modified. The results show:

  • Which conditions evaluated to TRUE or FALSE

  • What actions would be taken

  • Any configuration issues

Note: You must save changes before testing. The Test button is disabled if you have unsaved changes.


Step 7: Activate Your Playbook

Once satisfied with your configuration:

  1. Ensure all workflow steps are complete (no empty conditions)

  2. Click Save if you have unsaved changes

  3. Click Activate

The playbook will run automatically during scheduled collection cycles.

Deactivating a Playbook

To pause a playbook:

  1. Click Deactivate

  2. The playbook stops running but retains its configuration

  3. Click Activate again to resume


Troubleshooting

"Cannot activate playbook without steps" error

Cause: Your playbook has no configured workflow steps.

Fix: Add at least one condition or action step before activating.

Playbook not running on expected customers

Cause: Customer assignment conflicts or inactive status.

Fix:

  1. Verify the playbook is Active

  2. Check customer targeting includes the expected customers

  3. Remember customers can only be in one active playbook at a time

Condition branches have no available actions

Cause: All action types are already used in this branch.

Fix: Each action type can only appear once per branch. If you need multiple emails, consider restructuring with nested conditions.

AI conditions not evaluating correctly

Cause: Prompt may be ambiguous or not matching customer responses.

Fix:

  1. Make your AI prompt more specific

  2. Use the Test feature to see how conditions evaluate

  3. Consider using expression conditions for precise numeric thresholds

"Cannot save - incomplete steps" error

Cause: A condition has empty TRUE or FALSE branches with no actions.

Fix: Add at least one action to each branch, or delete the incomplete condition.


Best Practices

  1. Start simple: Begin with a single condition and two actions, then add complexity as needed

  2. Test thoroughly: Run tests with different customer scenarios before activating

  3. Use expression conditions for numbers: AI conditions work best for text analysis; use expression conditions for thresholds like days overdue or invoice amounts

  4. Document with descriptions: Use clear playbook names and descriptions so your team understands the workflow

  5. Review regularly: Adjust playbooks based on collection performance metrics