Support teams are under constant pressure to respond faster and handle more requests without adding headcount. Customers expect instant answers, whether they reach out at 2 PM or 2 AM. Traditional support models can't keep up with this demand, which is why more teams are turning to AI-powered tools.
HubSpot's customer agent is one of these tools. It's an AI assistant that works inside Service Hub to answer questions, resolve issues, and help customers without a human agent stepping in. The customer agent uses your knowledge base articles, website content, and other sources to give accurate, helpful responses in real time.
In this blog, we'll explain what a customer agent is, how it works in HubSpot Service Hub, and how it can help your team scale support while keeping customers happy.
A customer agent is an AI assistant powered by Breeze, HubSpot's AI platform. It lives inside Service Hub and handles support conversations across channels like live chat, email, WhatsApp, and Facebook Messenger.
The agent answers questions, performs actions, and escalates to a human when needed.
Unlike a chatbot that follows a scripted decision tree, a customer agent uses natural language processing to understand what customers are asking and responds with contextual, conversational answers.
It learns from the content you feed it, such as knowledge base articles, website pages, PDFs, and public URLs.
The customer agent detects the customer's browser language and responds in that language automatically. This makes it easier to support global customers without needing multilingual staff on standby around the clock.
AI chatbot adoption is growing fast. One-third of Americans claim to have used an AI chatbot in the past three months, showing how quickly people are getting comfortable with AI-powered support. For businesses, this means customers are more open to interacting with AI than ever before.
The customer agent works by connecting to your existing content and using AI to answer questions based on what it finds. Here's how the process works step by step:
When you create a customer agent, you choose what content it should use to answer questions. This can include HubSpot knowledge base articles, website pages, landing pages, blogs, uploaded files (like PDFs or Word docs), or external URLs.
The more relevant and up-to-date your content is, the better the agent performs. If you have a solid library of FAQs, help articles, and product documentation, the customer agent can pull from these sources to give accurate answers.
You can also create short answers manually. These are specific question-and-answer pairs you write yourself to cover common or high-priority questions that might not be fully addressed in your existing content.
When a customer sends a message, the agent uses natural language processing to understand what they're asking. It doesn't rely on exact keyword matches. Instead, it interprets the intent behind the question.
For example, a customer might ask, "How do I cancel my subscription?" or "What's the process for ending my account?" Even though the wording is different, the customer agent recognizes both as the same question and responds with the right information.
The customer agent evaluates how confident it is in its answer before responding. If it has high confidence, it gives a direct answer and includes a link to the source so the customer can read more if needed.
If the agent is less confident, it might ask follow-up questions to clarify what the customer needs. For example, if someone asks about pricing, the agent might ask which product or plan they're interested in before responding.
If the agent can't answer at all, it hands the conversation off to a human agent based on the rules you've set up. This ensures customers always get help, even if the AI doesn't have the answer.
The success rate for AI support is impressive. Customer service teams using AI have resolved issues 44% faster, proving that AI can handle support tasks efficiently when set up correctly.
Beyond answering questions, the customer agent can also perform actions through API integrations with your external apps. These actions are triggered by specific customer requests.
For example, if a customer asks to check their order status, the agent can make an API call to your order management system, retrieve the information, and respond with the status. If someone asks to reset their password, the agent can trigger that process automatically.
To set up actions, you define trigger phrases (like "check my order" or "reset password"), specify what information the agent needs to collect from the customer (like email address or order number), and configure the API endpoint the agent should call.
This takes repetitive tasks off your team's plate. Instead of manually looking up order statuses or resetting passwords for every customer who asks, the agent handles it instantly.
When the customer agent can't resolve an issue, it transfers the conversation to a human agent. You control how this handoff works.
You can set it to assign the conversation to specific users or teams, or leave it unassigned so your team can triage it. You can also configure handoff triggers based on specific words or phrases.
For example, if a customer mentions "refund" or "cancel," the agent can immediately route the conversation to a human.
This ensures complex or sensitive issues get the attention they need without the customer having to repeat themselves.
No, a customer agent is not the same as a support agent, but they work together.
Think of the customer agent as the first line of support. It resolves straightforward issues instantly, which frees up your human agents to focus on the cases that need more attention.
The two work best when they're used together. The customer agent deflects a large portion of incoming requests, and human agents step in only when necessary. This creates a faster, more efficient support operation without sacrificing quality.
Customer agents play several important roles inside HubSpot Service Hub. Here's how they fit into your support operation:
A large percentage of support tickets are repetitive. Customers ask the same questions about account setup, password resets, billing, shipping, and product features over and over.
The customer agent handles these questions automatically, which reduces the number of tickets your human agents need to touch. For high-performing teams, customer agents can resolve most conversations automatically.
This gives your support team more time to focus on complex issues, feature requests, and personalized customer interactions that actually require human judgment.
Customers don't only need help during business hours. They reach out late at night, on weekends, and during holidays. Without a customer agent, those requests sit in the queue until someone is available.
The customer agent works around the clock. It answers questions and resolves issues instantly, no matter when the customer reaches out. This improves response times and keeps customers from waiting hours or days for basic help.
Customer expectations around chatbots are improving, too. 80% of people say their chatbot interactions were generally positive, showing that customers are willing to engage with AI when it's done well.
As your business grows, support volume grows with it. Hiring more agents is expensive and takes time. The customer agent gives you a way to scale support instantly without increasing payroll.
You can handle more conversations with the same team size, which keeps costs down and helps you grow sustainably. Even small teams can offer enterprise-level support by letting the customer agent handle the bulk of routine inquiries.
When human agents spend less time on repetitive tasks, they can focus on higher-value work. They can spend more time coaching customers, solving complex problems, and identifying upsell or cross-sell opportunities.
This shift improves job satisfaction, too. Support agents often feel burned out when they answer the same questions all day. Letting the customer agent handle those tasks frees them up to do more meaningful work.
The Performance tab in Service Hub shows you how the customer agent is performing. You can see which questions it's answering, which ones it can't answer, and where knowledge gaps exist.
This data helps you improve your support content over time. If the agent keeps getting asked the same question it can't answer, that's a signal to create a knowledge base article or short answer for it. Over time, your customer agent gets smarter and more effective.
The customer agent can only answer questions based on what you give it, so take time to gather your help articles, FAQs, and product documentation before you begin. Once that's in place, the technical setup is simple.
Here's how to get started:
Before you create a customer agent, you need at least one connected channel. The agent works across live chat, WhatsApp, Facebook Messenger, email, and calling (in beta).
Make sure your preferred channel is already set up in HubSpot's inbox or help desk. If you're using live chat on a non-HubSpot website, add the HubSpot tracking code to your pages first.
The customer agent is only as good as the content you give it. Pull together your knowledge base articles, help documentation, FAQs, and product guides that answer common customer questions.
Take time to audit your content for accuracy and completeness. If your help articles are outdated or have gaps, the agent won't be able to provide good answers.
Go to Service Hub and navigate to Customer Agent. Click "Set up your agent" and give it a name.
Choose a personality for the agent: friendly, professional, casual, empathetic, or witty. If you've set up a brand voice in HubSpot, you can use that instead. This affects how the agent communicates with customers.
Click "Add existing HubSpot content" to select knowledge base articles, website pages, landing pages, or blogs. You can also upload files like PDFs or Word docs, or sync public URLs.
If you want the agent to crawl all related pages on a domain (up to 1,000 URLs), select the "Import related URLs" checkbox.
For questions not covered in your existing content, create short answers. These are manual question-and-answer pairs you write for high-priority or frequently asked questions.
Click "Assign" and choose the channel where you want the agent to work, such as live chat, email, WhatsApp, or Facebook Messenger.
Decide what happens if the agent can't answer a question. You can set it to hand off to a human agent (routing to specific users or teams) or stay assigned and send a configured message. Choose the option that fits your team's workflow.
Click "Test [agent name]" to open a preview window. Ask the agent common questions your customers would ask and try different phrasings to see how they respond.
Testing helps you catch content gaps before going live. If the agent struggles with certain questions, go back and add more content or create short answers. Testing doesn't use HubSpot Credits, so test as much as you need.
After testing, publish the agent and monitor its performance using the Performance tab. Track metrics like deflection rate, resolution rate, and customer feedback.
Check the Knowledge Gaps section regularly to see questions the agent couldn't answer. Create short answers directly from this page to resolve these issues and improve performance over time.
Customer agents represent a shift in how support teams operate. Instead of relying entirely on human agents to handle every request, teams can now use AI to resolve routine issues instantly while freeing up their staff to focus on complex, high-value interactions.
HubSpot's customer agent makes this possible by connecting to your existing content, understanding customer intent, and responding with accurate, conversational answers. It works 24/7, scales with your business, and continuously improves as you add more content.
For teams that want to deliver faster, more efficient support without sacrificing quality, the customer agent is a practical, powerful tool that fits directly into Service Hub.
Origin 63 helps businesses set up and optimize HubSpot's customer agent so it delivers real results. We'll help you organize your content, configure handoff rules, and train your team to use the agent effectively.
Ready to scale your support with AI? Contact Origin 63 today, and let's build a customer agent that works for your business.