HubSpot's Customer Agent: Latest Features and Updates Explained

Customer service teams are under increasing pressure to respond faster, resolve more requests, and deliver personalized experiences without continually adding resources.

AI has already changed how businesses approach that challenge. However, as customer expectations grow, answering common questions is only part of the equation. Customers also need help taking action, whether that's tracking an order, retrieving an invoice, or resolving an issue that would otherwise require a support representative.

That's where HubSpot's latest Customer Agent updates come in.

Throughout 2026, HubSpot has introduced a significant set of improvements that expand what Customer Agent can accomplish while making it easier for businesses to implement, manage, and optimize. From expanded actions and automated knowledge generation to more personalized support and improved testing, Customer Agent is becoming a more capable part of the customer experience.

Let's explore what's new, how the latest features work, and what these updates mean for teams looking to scale AI-powered customer support.

What Is HubSpot’s Customer Agent?

Customer Agent is an AI-powered agent within HubSpot's Breeze ecosystem that helps businesses automatically respond to customer inquiries, resolve common requests, and escalate issues requiring human assistance.

Using connected business knowledge, CRM context, and configured actions, Customer Agent can provide relevant answers and perform supported tasks across customer communication channels, including live chat, email, and messaging platforms.

Unlike traditional rule-based chatbots that rely on predetermined conversation paths, Customer Agent can interpret natural-language questions, retrieve relevant information, and determine how to respond based on its available knowledge and instructions.

The latest updates build on these foundations, giving businesses more ways to put Customer Agent to work across their service operations:

1. Expanded Actions Help Customer Agent Resolve More Requests

One of the most significant areas of improvement is Customer Agent's growing ability to take action on behalf of customers.

Customer Agent has supported actions through API connections for some time, allowing businesses to automate tasks such as checking order statuses or initiating password resets.

Recent updates make these capabilities easier to configure and expand the range of tasks the agent can perform through native HubSpot functionality and connected applications.

A growing library of prebuilt actions

HubSpot's expanded actions library provides a searchable collection of preconfigured actions that businesses can add to Customer Agent without building each interaction from scratch.

Depending on the action and configuration, Customer Agent can now:

• Retrieve customer and CRM record information.
• Create or update CRM records and properties.
• Enroll contacts or tickets in HubSpot workflows.
• Look up order information through supported integrations.
• Retrieve invoices and provide payment information.
• Execute custom API requests for additional business processes.

Previously, enabling many of these interactions required configuring API requests to connect Customer Agent with the appropriate systems.

The expanded library simplifies that process by giving teams predefined starting points for common actions. Businesses can still configure custom API actions when their requirements extend beyond the available options.

For service teams, this increases the range of customer requests that can be handled within a single conversation.

New payment and invoice capabilities

Customer Agent can also support more of the billing and purchasing process.

With the appropriate actions configured, the agent can retrieve invoices, provide details about charges, send invoice copies, and share secure payment options with customers.

For example, imagine a customer contacts support requesting a copy of an outstanding invoice.

Rather than requiring a representative to locate the record and send the information manually, Customer Agent can retrieve the relevant invoice and provide it within the conversation after the required verification.

HubSpot has also introduced payment link actions that allow Customer Agent to share an appropriate, preconfigured payment link when a buyer is ready to make a purchase.

These updates create opportunities to reduce administrative work and make billing-related interactions easier for customers to complete.

What this means for businesses: Customer Agent can contribute to more of the resolution process, helping customers complete routine requests while reducing the manual tasks that would otherwise occupy service representatives.

Note: Some payment and invoice actions require beta enrollment, appropriate payment configurations, and identity verification.

2. Faster Setup and More Effective Knowledge Management

An AI agent's effectiveness depends heavily on the information it can access.

Businesses need accurate, relevant knowledge to help Customer Agent understand customer questions and provide useful responses. But gathering, organizing, and maintaining that information can require considerable effort, particularly when support knowledge is spread across tickets, conversations, and internal documentation.

HubSpot's latest improvements make it easier to build that foundation using information businesses already have.

Guided Customer Agent setup

HubSpot has introduced an automated setup experience designed to help businesses get Customer Agent running faster.

Rather than configuring the agent entirely from scratch, the guided experience can draw on existing HubSpot data and historical conversations to help establish its initial configuration.

This includes recommendations for elements such as the agent's identity, tone, knowledge sources, and handoff behavior.

Teams can review these recommendations, adjust the configuration, and test the agent before deploying it to customer-facing channels.

This is particularly useful for businesses introducing AI-powered support for the first time or expanding into additional service use cases.

Generate knowledge from existing support conversations

Another notable improvement is Customer Agent's ability to generate knowledge from historical customer interactions.

Traditionally, organizations have relied on knowledge base articles, website content, and manually written answers to train customer-facing AI tools.

But some of the most valuable support information already exists in resolved tickets and previous customer conversations.

HubSpot's knowledge generator can analyze resolved tickets, inbox conversations, and uploaded files to identify recurring questions and create draft FAQ-style answers.

Teams can specify which information should be included, exclude particular topics, and review the generated knowledge before making it available to Customer Agent.

Consider a business that frequently receives questions about subscription changes, billing processes, or account setup.

Instead of manually reviewing hundreds of historical tickets to document how those questions were resolved, the team can use knowledge generation to identify common themes and create reusable answers.

The review process remains important. Generated content should be checked for accuracy, relevance, and consistency before it becomes part of the agent's knowledge.

Connect existing sources of customer knowledge

Customer Agent also continues to support a range of existing knowledge sources, including HubSpot website pages, landing pages, knowledge base articles, imported website URLs, and supported files.

Teams can combine these sources with generated knowledge and manually created short answers to give the agent a more complete understanding of their business.

The important consideration is maintaining the accuracy of that information over time. An agent trained on outdated policies, pricing, or procedures can still provide an incorrect response, regardless of how effectively it retrieves the underlying content.

What this means for businesses: Organizations can make use of existing customer knowledge with less manual preparation, accelerating implementation while building a more useful foundation for automated support.

3. More Personalized Support and Smarter Human Handoffs

Customer service rarely follows a single set of rules.

Different customers may have different products, subscription levels, support entitlements, or expectations. An answer that works for one customer might not apply to another.

HubSpot's recent Customer Agent updates give businesses more control over how the agent responds to different audiences and when human representatives should become involved.

Audience-specific customer experiences

With Customer Agent's new audience segmentation capabilities, businesses can tailor the agent's available knowledge to different customer groups.

Segments can be defined using information such as CRM properties, existing contact segments, and visitor context.

For example, a software company supporting multiple subscription tiers could configure Customer Agent to access different knowledge sources based on a customer's plan.

An enterprise customer might receive information about advanced functionality or dedicated support options, while a customer on a different plan receives resources relevant to their subscription.

Audience segments can also help organizations support customers in different regions or languages.

Rather than creating a separate agent for each audience, businesses can manage more tailored experiences within a shared Customer Agent configuration.

Note: Audience segmentation is currently identified as a private beta in HubSpot's September 2026 release documentation.

More control over agent behavior and escalation

Personalization is only one part of delivering an effective support experience. Businesses also need to determine which requests AI should handle and when human expertise is required.

HubSpot's updated guidelines give teams more flexibility to define how Customer Agent should behave in specific scenarios.

These instructions can establish response expectations, identify situations requiring escalation, and determine how the agent should proceed when it cannot resolve a request.

For example, a business might allow Customer Agent to answer standard billing questions while requiring a human representative to handle payment disputes or account-specific exceptions.

Clear guidelines help ensure the agent's behavior aligns with established service processes and customer expectations.

Improved routing and Help Desk visibility

HubSpot has also introduced improvements to how conversations move between Customer Agent and human service teams.

With fallback routing, businesses can direct eligible conversations to Customer Agent when human representatives are unavailable.

Teams can also assign conversations to Customer Agent directly from Help Desk or Inbox and use improved handoff visibility to understand which tickets still require human attention.

These capabilities help businesses incorporate AI into their existing service operations without losing visibility into customer requests.

What this means for businesses: Service teams can deliver more relevant customer experiences while maintaining control over escalation, workload distribution, and the interactions that require human involvement.

4. Improved Testing, Coaching, and Reporting

Deploying Customer Agent is only the beginning.

As businesses introduce more knowledge sources, actions, and customer scenarios, they need reliable ways to evaluate how the agent responds and identify opportunities for improvement.

HubSpot's latest testing and reporting updates make that process more accessible.

Test Customer Agent before deployment

One of the most useful improvements is batch testing.

Instead of testing individual questions one at a time, teams can create collections of questions and evaluate how Customer Agent responds across multiple scenarios.

These questions can be written manually or generated using topics from real customer conversations.

For example, a service team preparing to deploy Customer Agent could test a collection of common questions about account access, billing, subscriptions, and order tracking.

After running the test, the team can review the agent's responses, identify where adjustments are needed, and refine its knowledge or instructions.

Test collections can also be saved and reused, making it easier to evaluate whether changes improve the agent's responses over time.

HubSpot has also introduced an embedded tester that remains accessible throughout the configuration process, allowing teams to evaluate responses as they adjust the agent.

For businesses configuring custom API actions, additional testing capabilities help validate requests before publishing them.

Identify coaching opportunities

Testing helps businesses prepare Customer Agent for deployment, but live customer interactions provide additional insight into how well it's handling real requests.

HubSpot's coaching improvements help teams identify areas where Customer Agent may need additional knowledge, updated instructions, or better-configured actions.

For example, coaching opportunities might surface when the agent repeatedly struggles to answer questions about a particular topic, encounters conflicting knowledge sources, or cannot complete an expected action.

Rather than requiring teams to review conversations individually to discover recurring issues, these insights help prioritize where to make adjustments.

HubSpot also provides visibility into the knowledge sources and reasoning behind individual responses, helping teams understand why the agent handled an interaction a particular way.

Measure results and identify opportunities for improvement

Improved reporting and automatic topic organization provide additional visibility into Customer Agent's effectiveness.

Teams can analyze metrics such as resolution rates, human handoffs, conversation volume, customer feedback, and the topics customers most frequently ask about.

Topic-level insights help identify the categories of requests the agent handles successfully and those requiring additional attention.

It's also important to distinguish between deflection and resolution.

A conversation that ends without a human handoff does not necessarily mean the customer's issue was solved. Evaluating resolution alongside customer feedback, repeat inquiries, and escalation patterns provides a more useful picture of the experience.

What this means for businesses: Teams can take a more structured approach to optimizing Customer Agent, using actual customer interactions to improve response quality, expand supported use cases, and evaluate its contribution to service operations.

5. Customer Agent Expands Across More Communication Channels

Customers engage with businesses across a growing number of touchpoints, and their support preferences can vary depending on the issue.

HubSpot's recent updates expand how Customer Agent can assist customers across different channels while giving businesses more control over where it responds.

Customer Agent Voice

One of the most significant developments is the expansion of Customer Agent into voice-based support.

Customer Agent Voice, currently available in public beta, allows businesses to use the agent to answer incoming calls and respond to customer questions conversationally.

The agent can use connected knowledge and relevant context to assist callers, with the ability to transfer conversations to human representatives when needed.

This creates new opportunities for businesses handling high volumes of repetitive phone inquiries.

For example, Customer Agent Voice could help answer common questions about account processes, operating hours, or available services, reducing the number of routine calls that require a representative.

For service teams, voice expands the potential role of Customer Agent beyond digital messaging and into another important part of the customer experience.

More targeted email and form responses

HubSpot has also improved how Customer Agent responds to incoming emails and form submissions, giving businesses more control over which interactions receive automated support.

New email targeting capabilities allow teams to use existing HubSpot lists and filters to specify which contacts Customer Agent should respond to. For example, a business could enable automated email support for customers participating in a product trial while maintaining a separate support process for accounts with dedicated representatives.

Customer Agent can also respond directly to eligible form submissions. Rather than sending a generic confirmation email, the agent can use the information submitted and its connected knowledge to generate a relevant response.

For example, if someone submits a website form asking how to configure a particular feature, Customer Agent can provide instructions or direct them to the appropriate resources without requiring a representative to respond manually.

These features build on the agent's existing support for live chat, WhatsApp, and Facebook Messenger.

What this means for businesses: Organizations can extend AI-powered assistance across more customer touchpoints while retaining greater control over which conversations are handled automatically.

How to Make the Most of HubSpot's Latest Customer Agent Updates

With more capabilities available, businesses have an opportunity to expand how they use Customer Agent across their service operations.

But realizing the value of these updates requires a thoughtful approach to implementation.

Start with the customer requests that create the most manual work

Before introducing additional actions or channels, evaluate where your service team spends the most time.

Look for high-volume, repeatable requests with clear resolution criteria.

Common examples include account questions, invoice retrieval, order tracking, and other routine inquiries that follow established processes.

These are often practical starting points for Customer Agent because businesses can define the information, actions, and conditions needed to resolve them.

Connect the knowledge, data, and workflows behind each use case

As Customer Agent takes on more responsibility, the systems supporting it become increasingly important.

An agent retrieving customer information needs access to accurate CRM data. An agent performing actions needs correctly configured workflows and integrations. An agent escalating a request needs routing rules that connect customers with the appropriate team.

Before expanding its responsibilities, evaluate whether those foundations can support the interactions you want the agent to handle.

For each use case, identify the knowledge required, the systems involved, the action the agent should take, and the circumstances that require human intervention.

Use deployment recommendations to prioritize where to launch

HubSpot's deployment recommendations can help teams identify promising opportunities using their existing service data.

These recommendations consider factors such as conversation volume, response times, after-hours activity, unassigned conversations, and service-level agreement adherence.

Teams can also use Breeze Assistant to explore these recommendations and better understand where Customer Agent could help.

This creates a more informed starting point for choosing channels and determining how much conversation volume to assign to the agent.

Treat optimization as an ongoing process

Customer needs, business processes, and support knowledge will continue to evolve after deployment.

Build a regular process for reviewing Customer Agent's results, updating knowledge sources, evaluating handoffs, and testing additional scenarios.

As the agent demonstrates reliable results in its initial use cases, businesses can expand into additional actions, audiences, and communication channels.

The opportunity is to develop a customer support operation where AI and human representatives work from connected information, follow established processes, and contribute to a consistent customer experience.

Frequently Asked Questions About HubSpot Customer Agent

How much does HubSpot Customer Agent cost?

Customer Agent uses HubSpot's credit-based pricing model.

HubSpot currently lists messaging at $0.50 per resolution and voice at $0.50 per minute.

For messaging, a qualifying resolution generally involves the agent providing support without a qualifying handoff to a human representative during HubSpot's 72-hour evaluation period. HubSpot also defines certain lead qualification outcomes as resolutions.

Organizations should review current credit allowances, usage limits, and eligibility requirements when estimating costs.

Does Customer Agent replace human support representatives?

Customer Agent is designed to handle eligible customer inquiries and routine support tasks while working alongside human service teams.

Businesses can configure guidelines and handoff processes to direct customers to representatives when additional expertise, judgment, or intervention is required.

Its value comes from helping teams handle more routine requests while preserving human involvement where it matters most.

Are all the latest Customer Agent features available now?

Availability varies depending on the feature, HubSpot subscription, and release stage.

Some capabilities, including Customer Agent Voice and audience segmentation, have beta-specific availability. Certain actions may also require additional configuration or beta enrollment.

Businesses should confirm access and requirements within their HubSpot portal before planning an implementation.

The Next Stage of AI-Powered Customer Service With HubSpot

HubSpot's latest Customer Agent updates represent a meaningful expansion of what's possible with AI-powered customer support.

With more actions, easier knowledge management, personalized experiences, additional channels, and improved optimization tools, businesses have more opportunities to resolve customer requests efficiently and scale their service operations.

For organizations already using HubSpot, these developments create new ways to get more value from their existing platform. For teams just beginning to explore AI-powered service, they provide a more accessible starting point for automation.

The biggest opportunity is putting these capabilities to work within a connected customer experience, supported by reliable data, effective workflows, and clear service processes.

Explore New Breed's AI Services to discover how we can help you put HubSpot's latest capabilities to work for your business.