HubSpot AI Agents: What results to expect?

Would it be cost-effective for your team and company to invest in HubSpot AI agents? Several factors can influence the degree of success of AI agents, but if you’re already using HubSpot, there’s a good chance that the short answer is yes.

As an example, here are some results reported by HubSpot for each of its three main agents:

• Customer Agent can resolve 72% of tickets without human intervention (second quarter 2026), compared to about 20% a year earlier.
• Prospecting Agent enables on average 82% more meetings with prospects, and 76% more qualified leads.
• Content Agent, which is still evolving, currently shows an average 14% increase in reach.

Let's explore the potential of each of these AI agents.

Prospecting Agent:  more conversations, less forgetting

HubSpot's Prospecting Agent monitors external signals (new hires, rounds of funding, mergers, acquisitions, etc.) and internal signals (website visits, returning prospects, e-mail opens, etc.) to help your prioritize your prospecting activities, write e-mails and establish an approach strategy. Here's what to expect:

• More conversations and meetings
  

  The agent can propose meetings and handle a larger volume than a single person. We see it working for our clients every day.
  
  

• No missed follow-ups

  The agent always sends a follow-up at the right time, or offers to do so, which creates a significant effect on sales. A great starting point is reviving dormant opportunities.
  
  

• Large-scale hyperpersonalization

  Being able to research each prospect personally would be ideal, yet it's not viable at medium or large scale. The agent can handle that for you.
  
  

• Time saved

  This is our favorite indicator! Someone can finish a little earlier on Friday because they reached their target for prospects to contact. Leaders who have lost team members and are hesitant to hire can actually use this agent to do more with less.
  

   

Real example: Jotform generated 26 qualified deals on the 1,000 tested contacts, very close to their representative's actual results.

Customer Agent:  faster answers and a more available team

The agent answers requests that come in via chat, e-mail or even phone using your company's documentation and processes, kind of like a new employee with access to a knowledge base. The results can be seen on three main levels: 

• A majority of requests processed automatically

  Current usage statistics indicate 70-90% of requests are fully processed by the agent, depending on the organization. Complex or sensitive requests are sent to the team.
  
  

• Happier clients

  For order tracking, for example, the customer doesn't need to wait for a person to be available to find out where their package is. They can receive an immediate answer for more administrative cases where human conversation has little added value.
  
  

• Sales (you guessed it!)

  Since the agent is chatting with people with questions that may intend to purchase, it can provide them with the information they need and even suggest a meeting with a sales representation. Overall, depending on the company, this agent can become an additional channel for qualified lead generation.

Real example: 60% of Sesame HR tickets are resolved automatically, without escalation.

Content Agent:  branded content and better SEO

Content Agent drafts articles, LinkedIn posts and landing pages directly in HubSpot based on your brand tone, account data and AEO recommendations. The documented results are fewer than the other two agents for now, but here is what we're seeing:

• The end of the blank page

  The agent gives you base content using your brand tone settings and data from your HubSpot account. It's the same logic as the other agents: the more carefully those settings are set, the closer the draft is to on-brand .
  
  

• Better visibility in search and answer engines

  HubSpot's AEO tool queries AI engines daily with the prompts you set to measure your brand visibility. The recommendations it gives can be a great starting point for a blog post or public LinkedIn content.
  
  

• Increased reach

  Although the method and time period are not specified, HubSpot has mentioned seeing a 14% higher reach on average.

What does it take to see results with HubSpot's AI agents?

We're loving HubSpot's AI agents, but realistically, they don't deliver results at the touch of a button. You won't suddently have the equivalent of 10 new team members contacting everyone and generating tons of new appointments automatically.

The reality is more like the arrival of a new team member. They need to be trained, checked on, and gradually trusted. We've noticed that these four factors make the biggest difference:

• CRM data

  The agent uses your CRM context. If your team doesn't use the CRM or have their inboxes and calendars connected, clearly the results will be affected. 
  
  

• A specific use-case

  "Follow-up with my prospects" is too vague. Which prospects? The ones we haven't reached out to in six months? What industry are we targeting? And is the goal to book a meeting, make an announcement, invite to a webinar? This is where a strategy set by an agency like Guépard can make all the difference.
  
  

• A responsible human

  All our clients to date use Prospecting Agent in semi-autonomous mode, meaning that someone reviews e-mails before they're sent. The agent learns from your corrections, like a new employee being coached. This same principle applies to all HubSpot AI agents. 
  
  

• Quality over quantity

  Aiming for high quantity is tempting but risky. Users who don't see the results they're hoping for typically moved too fast instead of going step by step.

Traps that  limit results

These agents aren't necessarily complicated to deploy, and they have huge potential, but we see the same mistakes over and over. Here's what to look out for:

• Using someone else's personalization
  

  Say your salespeople are all assigned a specific region. If a salesperson uses his colleague's agent, who uses she/her pronouns, the agent may reach out to people from her region instead of his. It also may write e-mails using the incorrect pronouns.
  
  

• Not integrating the agent into the team's routine

  BDRs and/or sales need to know where to find the prepared e-mails and review them every morning as they plan their daily tasks. It can seem time-consuming at first, but trust us: the investment will be worth it by reducing follow-ups and post-mortems.
  
  

• Forgetting the contact's preferred language
  

  This may seem silly, but if the option isn't activated in the configuration, e-mails could be sent in the default language if they are not carefully checked.
  
  

• Automating what should stay human

  Some needs can be met by an agent, but others should be met by a person. It is important to distinguish cases and be sensitive to process complicance.
  
  

• Ignoring brand tone

  Brand tone can be added to HubSpot and/or agent configuration. Our advice is to avoid giving the agent a humorous tone, as it can be perceived as being sarcastic or arrogant in certain contexts.

How to mesure the AI agent's results?

We recommend that you document your starting point with data and facts prior to launching an AI agent. Without that, you may have a hard time figuring out what's changed. Here are some examples or KPIs you can monitor:

Results harder to quantify that bring real impact

In overworked and labor-strapped teams, HubSpot’s AI agents can be a game changer or even a mood changer for the better.

Let’s take the example of one of our clients: a salesperson listened to voice recordings of evaluations on site, transcribed them and then wrote a quote. This is exactly the kind of redundant task that can discourage a team member in time. When artificial intelligence (AI) removes these tasks, people are more inclined to stay and continue evolving in the company.

In the past few months, we've also gotten to see people start falling in love with their work again. This springs from the innovation of AI requiring their knowledge, expertise and experience. There's a playful side to all of this that we can't ignore.

How Guépard supports the deployment of HubSpot's  AI agents

By collaborating with Guépard, a HubSpot Platinum agency in Québec, you could avoid errors that we see often and therefore deploy your agents more effectively, thus seeing results faster. It's a bit like exploring a forest with a guide: you can do it alone, but if you go with someone who knows the terrain, they'll know which paths are the best and the most slippery.

Our method consists of a few steps: 

1. A first meeting to talk about strategy, segmentation and goals.

2. An audit of CRM data (if we see issues, we'll either tidy before launching or change the strategy).

3. A launch on a small sample to be able to refine the message, the target and the interactions with the team.

4. Adjustments and, if needed, a training for your team to become self-sufficient.

Still have questions on the possible results HubSpot's AI agents can bring you? We're here to help!