TG Bot Multi-Agent Routing Guide: Online Priority, Designated Agents, and Project Scope Explained

When operating a Telegram Bot community or customer service team, as user messages grow to hundreds or even thousands per day, relying on a single admin account to reply in turn quickly leads to “can’t keep up, lost conversations, dissatisfied users.” The core value of TG Bot multi-agent customer service is to enable multiple agents to handle conversations in parallel, distributing workload like a professional call center—no missed leads, seamless collaboration.

TG-Staff, as a customer service and operations SaaS platform for Telegram Bots, offers out-of-the-box multi-agent routing. This article will help you understand the logic behind routing, guide you through configuration step by step, and provide advanced tips and common troubleshooting solutions.

---

Why Does TG Bot Customer Service Need Multi-Agent Routing?

Single Agent vs. Multiple Agents: Where’s the Efficiency Gap?

Suppose you run a cross-border e-commerce Telegram customer service Bot:

• Single-agent mode: User A asks → agent replies → User B waits → agent finishes with A before replying to B. If the agent needs to look up information, take a meal break, or finish work, User B might wait 30+ minutes or even churn.
• Multi-agent routing mode: Users A and B message simultaneously; the system automatically assigns them to two different online agents who handle them in parallel. Average response time can drop from 10 minutes to under 1 minute.

Based on real operational experience, when daily conversations exceed 50, single-agent reply latency significantly lowers user satisfaction, making multi-agent routing almost essential.

How Does Routing Rules Affect Customer Service Conversion?

Improper routing rules can backfire. For example:

• Fully random assignment: May cause the user to be handled by different agents each time, requiring them to repeat their issue—terrible experience.
• No online detection: Messages assigned to an offline agent never get a reply.

The correct routing rules should ensure: Messages reach an authorized, online agent first, and the same conversation is always handled by the same agent (unless actively transferred). TG-Staff’s two routing modes are designed for this.

---

Core Concepts of Multi-Agent Routing: Project Scope and Agent Permissions

Before configuring routing, you need to understand two basic settings in TG-Staff: Project customer service scope and Agent permissions.

Project Customer Service Scope: All Agents vs. Designated Agents

In the TG-Staff console, each Bot project can set its “customer service scope”:

Mode	Description	Use Case
All Agents	All agents (regardless of permissions) can see and handle conversations for this Bot	Single-brand Bot, whole team needs to handle chats
Designated Agents	Only selected agents can see conversations for this Bot	Multi-brand Bot (e.g., company runs Bot A and Bot B), each Bot assigned to specific team members

For example, if you run both a “Pre-sales Bot” and a “Post-sales Bot,” you can assign the pre-sales team to the former and the post-sales team to the latter, preventing agents from receiving irrelevant conversations.

Agent Permissions and Operation Scope Configuration

In “Agent Management,” you can set which projects each agent can manage. Admin-level agents can view all conversations across projects, while regular agents only see their authorized projects.

Configuration path: 控制台 → 坐席管理 → 编辑坐席 → 项目权限.

---

Detailed Explanation of Two Distribution Modes: Round Robin vs Online Priority

TG-Staff offers two distribution algorithms, which can be switched in Project Settings → Session Distribution.

Round Robin

• Principle: The system maintains a round-robin list of agents and assigns new sessions to agents in a fixed order (e.g., Agent A → B → C → A → …).
• Advantages: Fair distribution, suitable for teams with fixed schedules and a small number of agents (3–5 people).
• Disadvantages: If Agent A is offline, messages still attempt to be assigned to A until a timeout occurs, then transfer to B, causing user wait times to increase.

Online Priority

• Principle: When a new session arrives, the system prioritizes assigning it to agents whose current status is “online” and who have permissions. If all agents are offline, it falls back to round-robin distribution (queued in round-robin order until someone comes online).
• Advantages: Messages are immediately handled by online agents, suitable for flexible schedules and cross-timezone teams.
• Disadvantages: If all agents are online for a long time, distribution may favor active agents (can be mitigated by increasing agent quota).

Recommendation: Most teams should choose “Online Priority” first. If you have fixed schedules and a stable number of agents, you can use “Round Robin” to ensure fairness.

How to Switch Distribution Mode

1. Log in to the TG-Staff console and navigate to the target Bot project.
2. In the left menu, select “Project Settings” → “Session Distribution”.
3. In the “Distribution Mode” dropdown, choose “Round Robin” or “Online Priority”.
4. Click Save.

---

Step-by-Step Guide: Configuring Multi-Agent Distribution for Your Telegram Bot

The following steps assume you have already registered for TG-Staff and added a Bot project.

1. Log in to the console: Open https://app.tg-staff.com and enter the target Bot project.
2. Create agent accounts: In “Agent Management”, click “Add Agent”, enter the email and password (or add via Bot invitation link). Depending on your plan, the Standard plan supports 3 agents, and the Pro plan supports 20.
3. Configure project agent scope: Go to “Project Settings” → “Agent Scope”, select “All Agents” or “Specified Agents”. If you choose “Specified Agents”, check the agents allowed to handle this Bot.
4. Select distribution mode: In “Session Distribution”, choose “Online Priority” or “Round Robin”, then save.
5. Test distribution: Send a message to the Bot using another Telegram account. Observe the console’s “Session List” — you should see the new session appear in one agent’s session list. Send multiple messages in a row to verify that distribution follows the rules.

---

Advanced Tips: Combining Diversion Links for Traffic Attribution

If you are promoting your Bot via ads or social media, you can leverage TG-Staff’s Diversion Link feature to track each visitor’s source and view the “Source” field in the user profile on the agent side.

How Diversion Links Work

A diversion link is a short URL on TG-Staff’s official domain (e.g., https://app.tg-staff.com/abc123). When a user clicks the link:

1. The system captures the visitor’s IP address, browser information (User-Agent), and custom parameters carried in the URL (e.g., utm_source=google_ads).
2. The page automatically redirects to the specified Telegram Bot chat window.
3. After the user sends a message, this information is written into the user profile, which agents can view in the conversation interface.

Best Practices

• Generate separate diversion links for different channels: e.g., one link for Google Ads, one for Twitter, one for Telegram groups.
• Append custom parameters to the link, such as ?campaign_id=summer_sale, for easier later analysis.
• Ensure the Bot pointed to by the diversion link has session diversion configured; otherwise, users may be directly assigned to the default agent.

---

Common Distribution Issues Troubleshooting and Optimization Tips

Issue 1: Agent is Online but Not Receiving New Conversations

• Check Scope: Is the agent authorized to handle this Bot? Verify in “Project Settings” → “Agent Scope”.
• Check Distribution Mode: If using “Round Robin”, confirm the agent is in the polling list (all authorized agents participate).
• Check Agent Status: Is the agent truly online? A green dot next to the avatar in the web portal indicates online status.

Issue 2: User is Repeatedly Assigned to Multiple Agents

• Cause: Usually due to “Conversation Transfer” being enabled and old conversations not being properly closed. When Agent A transfers a conversation to Agent B, if A does not close the session, the system may think A is still handling it, leading to multiple sessions for the same user.
• Solution: In “Conversation Transfer” settings, ensure “Auto-close original conversation after transfer” is checked. Also train agents: do not reply to the original conversation after transferring.

Optimization Tips

• Peak Hours: Use “Online First” + temporarily increase agent capacity (e.g., upgrade from 3 to 5). TG-Staff supports periodic plan upgrades; see the official pricing page.
• Fixed Shifts: Use “Round Robin” to avoid grabbing. Also set offline auto-reply messages to inform users “We will reply within X hours”.
• When All Agents Are Offline: User messages will enter the Bot’s auto-reply flow (e.g., welcome message/menu) and won’t be lost. Once an agent comes online, the system will reassign the conversation according to distribution rules.

---

Frequently Asked Questions

Q: Does the TG bot customer service multi-agent routing require an additional fee?
A: The session routing feature of TG-Staff is supported in the standard version at no extra cost. The standard version supports 3 seats, while the Pro version supports 20 seats. If your team is larger, please contact customer service to upgrade your plan.

Q: Which is more suitable for my team: online priority or round-robin assignment?
A: If your customer service team has fixed shifts and a small number of agents, round-robin assignment is fairer. If your team has flexible shifts or operates across time zones, online priority ensures messages are handled by online agents first. We recommend trying online priority first and then adjusting based on actual feedback.

Q: Can I assign an agent to handle messages from only a specific bot?
A: Yes. In the project settings, set the project agent scope to “Designated Agents,” then check the agents allowed to handle that bot. Unchecked agents will not see conversations from that bot.

Q: What user information can the routing links track?
A: Routing links can capture visitor IP, browser type (User-Agent), and custom parameters carried in the URL (e.g., utm_source, campaign_id). This information is written to the user profile and can be viewed by agents in the conversation interface.

Q: Will user messages be lost if all agents are offline?
A: No. User messages will enter the bot’s auto-reply flow (e.g., welcome message/menu) and be stored in the system. Once an agent comes online, the session will be reassigned according to routing rules. It is recommended to configure offline auto-reply messages to inform users of the wait time.

---

Experience TG-Staff Multi-Agent Routing Now

• Free trial for 3 days, no credit card required: https://app.tg-staff.com
• View full documentation: https://docs.tg-staff.com
• Join the official customer service group for inquiries: Contact @tgstaff_robot

Master TG bot customer service multi-agent routing configuration to let your customer service team efficiently handle every conversation without missing any potential users.