Skip to main content

Playbooks

How Stuut uses Playbooks to learn your business and run collections your way.
A Playbook is Stuut’s way of capturing how your business operates — your tone, your escalation rules, your exceptions, and your preferences — so the agent can run your collections workflow without constant supervision.A Playbook is made up of articles: short pieces of knowledge such as policies, SOPs, and process notes. There are two kinds of content:Knowledge — the facts about your business. Who your key contacts are, which customers get special treatment, what your standard payment terms are, and any account-specific notes your team has logged.Strategy — the rules governing how the agent behaves. When to send reminders, what tone to use, when to escalate, which accounts to treat differently, and what actions require human approval before the agent takes them.
Playbooks currently apply to the Collections module, including disputes and payment promises. The agent uses the Playbook to determine how to work each account — which cadence to follow, what tone to use, when to escalate, and when to pause outreach.Cash Application has its own settings under Playbook → Cash App, such as automatic adjustments. See Cash Application.
The agent checks your Playbook when it:
  • Drafts email replies
  • Decides whether to pause or resume outreach based on what a customer said
  • Classifies disputes and spots when a dispute is resolved
  • Decides whether a message needs a reply
  • Captures payment promises, applying rules like “promises over $10,000 need an acknowledgment email”
Articles are guidance only. Approving an article never sends anything by itself, doesn’t override your workflows, and doesn’t bypass human review.
Yes. Tasks show citations to the Playbook articles the agent relied on, so you can see why it did what it did.
Access depends on your role. Admins can view and edit the Playbook and approve changes that affect how the agent behaves. If you can’t see Playbooks, ask your admin.
When you onboard to Stuut, your CSM works with your team to configure the starting Playbook. This involves:
  • Mapping your customer segments and any accounts that need special handling
  • Setting your collections cadence (timing, channels, escalation thresholds)
  • Defining your tone and communication style
  • Establishing approval rules — which actions the agent can take autonomously and which require a human sign-off
The Playbook is editable at any time after onboarding. You don’t need to get it perfect on day one.
Yes. Articles can be scoped three ways:
  • Shared (the default) — applies across your organization
  • Customer — only used when the agent is working that customer. This is good for payment preferences, portal rules, and key contacts.
  • Segment — used for every customer in a segment, so one article can cover many customers
Not automatically. Playbook articles are the guidance your team writes and approves, and they stay as written until you change them. Automatic suggested updates based on what the agent sees in new conversations are on the roadmap.If you see older items in a review queue from your onboarding, they may be out of date. Review them before approving.
  • Be concrete and self-contained. Each article should make sense on its own.
  • Cover every branch. If the rule is “if X, do A; otherwise do B,” include both.
  • Make it actionable. Say what the agent should do, not just background.
  • Avoid contradictions. Two articles that disagree lead to unpredictable behavior, so consolidate duplicates.
  • Scope it. If an article is about one customer, scope it to that customer.
No. Create articles individually. If you have a long SOP, break it into focused articles, one topic each.
The Playbook is accessible from the main navigation in Stuut under Playbooks. From there you can:
  • View the full Playbook for Collections
  • Add, edit, and archive articles
  • Edit strategy rules directly
  • Add or update account-specific knowledge
  • Manage segments and custom template tables
Changes take effect immediately — the agent uses updated articles the next time it handles a relevant task.