While the BusinessChat inbox excels at handling real-time customer conversations, important follow-up work often flows out of them. Whether it’s a personal callback, a cross-team escalation that takes days, or a widespread system outage, Work Items ensures no task slips through the cracks. By moving work out of loose chats and into a structured tracking system, your team maintains full context, managers get clear operational visibility, and customers experience seamless communication.
The Core Concept: One Umbrella, Three Types
BusinessChat uses a single, unified engine called Work Items (a culturally familiar term for tracked cases moving through a workflow).
A Work Item is simply the core object you interact with, and its Category is a property you choose when you create it. There are three categories tailored to different operational needs:
Tasks: Lightweight, personal follow-ups owned by a single user and driven by a due date (e.g., "Call the customer back Tuesday about their exchange"). Tasks are always tied to one customer contact.
Tickets: The classic support ticket for cross-team or departmental escalations governed by a Service Level Agreement (SLA) clock (e.g., "Refund not received—needs Finance to investigate"). Tickets are always tied to one customer contact.
Incidents: Top-level items used to track major, systemic issues affecting many customers at once (e.g., "Payment gateway outage" or "Carrier delay affecting 300 orders"). Incidents utilize a many-to-many relationship, letting you link multiple affected contacts to a single issue.
Work Items Types live under Work items Settings, you can distinguish which is a ticket, a task or an incident from the Type field. You can also, edit any work item's configuration by clicking the Pen icon next to it's name.
Why Work Item "Types" Matter
Before creating items, admins set up structured Work Item Types (such as a "Returns" ticket or a "Refunds" ticket) under each category.
By defining a Type first, you eliminate the chaos of free-form tracking. A configured Type automatically bakes the right custom fields, standard statuses, SLA targets, and routing rules directly into the workflow from second one. This ensures your entire team speaks the same operational language and reports compare apples to apples.
Configuration Scope Matrix
Each category behaves as an independent silo when it comes to standard workspace operations:
SLA Configuration: Supported strictly for Tickets and Incidents only; Tasks do not feature or track SLAs.
Automated Customer Notifications: Supported strictly for Tickets only; Tasks and Incidents do not trigger individual automated transactional alerts.
The Unified Lifecycle & The SLA Clock
Every single work item moves through a unified, five-stage lifecycle. While businesses can customize individual sub-statuses underneath these stages to match their exact workflows, the five parent stages are fixed so the system can accurately track metrics and time:
Submitted: Created but not yet picked up. The Time-to-Resolve (TTR) SLA clock starts here.
In Progress: Actively being worked on by your team. The SLA clock continues running.
Waiting on Customer: Paused on a customer dependency. This is the only stage where the SLA clock pauses, meaning it won't count against your team's performance. (Note: Incidents do not have a "Waiting on Customer" stage).
Resolved: A terminal state indicating the work is complete. Customer Satisfaction (CSAT) surveys are triggered here.
Canceled: A terminal state indicating the work has been called off.
⚠️ The SLA Trap to Avoid: Sub-statuses like "Waiting on 3rd Party" or "Waiting on Internal" sit under the In Progress stage. Because your company still owns the accountability of the work, the SLA clock keeps running!
Reopening Items
If a resolved or canceled item needs to be revisited, users simply move the status back to In Progress. The system automatically treats this as a Reopen, tracking your total reopen counts and resuming the exact same SLA clock right where it left off using the original item ID.
Smart Assignment & Auto-Routing
Work Items are designed to find the right person automatically and maintain context through two core routing behaviors:
Round-Robin: The very first time an item enters a specific status, it is automatically distributed fairly among the active members of that status’s owning team to prevent anyone from being overloaded.
Sticky Assignment: If an item leaves and later returns to a previously visited status (for example, after a customer replies to a pending question), it automatically routes back to the same team member who handled it last. This preserves user continuity and prevents agents from having to pass context back and forth.
Team-Level Ownership Constraints: Individual or "personal" back-pocket todos belong elsewhere; every single work item in the database must be structurally owned by an accountable backend team. Furthermore, once captured at the moment of creation, the ticket's original Creator Team remains entirely immutable, ensuring long-term tracking accountability even if the customer's live chat session later re-routes to a different department.
Note: Tasks do not support automatic routing rules and are always assigned manually.
Internal Team Collaboration
Every work item features a single, consolidated Internal Discussion Feed at the bottom of its details page. Internal comments from team members (complete with @mention and @team support) interleave chronologically with recorded system events (like status updates or assignment changes) and references to customer-facing messages.
This layout allows your team to understand full context and collaborate natively without constantly hopping back and forth between an activity tab and a comment view. Most importantly, this feed is strictly internal and is never visible to the end customer.
Feed Modification Guardrails:
To ensure an unalterable history of operations and handoffs, editing or deleting comments and recorded system events is entirely unsupported. The conversation and system history feed remains an uneditable, sequential log. Furthermore, comment threading is completely flat; nested or branched replies are disabled, and all text formatting remains simple to keep execution identical across desktop and mobile versions.
Summary: 5 Things to Remember
Everything is a Work Item: Tasks, Tickets, and Incidents share the same underlying machinery.
Categorize with Two Questions: Ask yourself: How many customers are affected? and Who needs to act?
Types Automate the Setup: Fields, statuses, SLAs, and routing are baked in automatically by type configuration.
One Clock, One Pause State: The global Time-to-Resolve clock pauses only when an item sits in the "Waiting on Customer" stage.
Round-Robin, Then Sticky: Work is distributed fairly first, then automatically routed for agent continuity.