Customer Success Perspective: Using AI to Create More Time for What Matters

How HubSpot CS Is Using AI to Do More of the Work That Matters

Reclaiming Time for Human Connection in Customer Success

Mark Henry, Senior Manager of Corporate Customer Success at HubSpot, shared a candid and practical look at how his organisation is deploying AI to fundamentally reshape customer-facing work. With five years at HubSpot and a career spanning industries from FMCG to fintech, Mark grounded his talk in the reality that AI adoption is not just a technology project — it is an exercise in change management, cultural transformation, and leadership conviction. HubSpot CEO Yamini Rangan's framing of "radical uncertainty" and the view that "inaction is terminal" set the tone for a team-wide push to embrace AI across every layer of customer success.

The central thesis of the session was that AI should absorb the mundane, repetitive, and admin-heavy tasks so that customer success managers can spend more time doing what actually drives value: having meaningful, consultative conversations with customers. Mark walked through three live success stories — the Breeze Customer Success Assistant, the Outreach Agent, and the Customer Support Agent — each illustrating how HubSpot's CS team is reclaiming minutes per interaction and scaling personalised engagement without adding headcount. The talk resonated well beyond CS teams, offering transferable lessons for anyone in a customer-facing role navigating the current wave of AI disruption.

The Macro Moment: Fear, Urgency, and the Case for Action

Mark opened by addressing the pervasive anxiety around AI — the fear that it could replace jobs, the pressure to adopt tools without fully understanding them, and the overwhelming pace of change. He referenced HubSpot's CEO Yamini Rangan, who speaks of "radical uncertainty" and "massive transformations," using that leadership conviction to drive the entire organisation toward action rather than paralysis.

• Customers are both excited and overwhelmed by AI and the pace of change, seeking more strategic, consultative guidance from their CSMs.
• HubSpot SVP Anne-Marie Costigan emphasised that the CS team's unique value proposition in an AI-dominated world is human connection.
• The challenge is finding the right mix between AI efficiency and genuine human engagement — a theme echoed throughout the session.

How CS Itself Is Changing: Shifting Expectations and a New CSM Identity

Mark described how the customer success role feels fundamentally different, recalibrating roughly every six months. What was once an adoption-management function now requires change-management skills, AI fluency, and even a sales-like ability to guide customers through discovery and experimentation.

• The CSM role has shifted from product-usage conversations to strategic AI-adoption conversations, requiring discovery questions, pain-point analysis, and consultative selling skills.
• Mark is now hiring people with sales backgrounds for CS roles because the ability to persuade and guide customers through change is essential.
• There is a genuine CSM identity challenge — the role's definition changes constantly, and teams must continuously retrain and recalibrate.
• Customers who are not getting outcomes and value — not just usage — will leave faster than ever, because AI has lowered the cost and speed of switching platforms.
• CSMs themselves must be AI-curious and hands-on with the product to maintain credibility with customers.

Internal AI Enablement: Building the Culture and Infrastructure

Before deploying customer-facing AI tools, HubSpot invested heavily in internal change management, experimentation, and governance. Mark outlined the infrastructure his teams have built to drive adoption and ensure strategic alignment from leadership through to frontline CSMs.

• The team actively trials tools like Perplexity, Claude, ChatGPT, NotebookLM, and Glean, working toward a consolidated AI tech stack suited to their workflows.
• AI Fluency Fridays — dedicated enablement sessions held on Fridays when customer engagement is lower — have been highly effective for coaching and skill-building.
• A Global AI Council brings together leadership, managers, and frontline CSMs to set strategic direction and build go-to-market plans collaboratively, ensuring buy-in from the people who execute.
• Close partnerships with sales tech and sales operations teams are critical; the CS org now benefits from operational support it previously lacked.
• Dedicated Slack channels for AI ideas and questions foster a culture of continuous sharing and experimentation.

The CS AI Roadmap: Giving Time Back at Every Stage of a Call

Mark presented an ambitious four-quadrant roadmap designed to remove low-value work from every phase of a CSM's customer interaction — from pre-call preparation and live call support to post-call wrap-up and outreach. The goal: push monthly call targets from 57 to 80 without burning out the team.

• Pre-call: An agent automatically surfaces last engagements, contact history, and health signals — no manual research required.
• During the call: A live assistant sits alongside the CSM, providing answers, links, and supporting details in real time when unexpected questions arise.
• Post-call: Meeting notes are generated from the call transcript and presented to the CSM for one-click sending, eliminating manual wrap-up.
• Outreach: An agent drafts personalised emails based on prior conversations, enabling CSMs to fill their calendars without manual email composition.
• Only two of these agents are fully live today (marked in orange on the roadmap), but the trajectory is clear and ambitious.

Success Story 1: Breeze Customer Success Assistant

The Breeze Assistant — accessible via a small widget icon on any CRM page — has become an indispensable part of daily CSM workflows. It evolved from a basic Slack bot that dumped account information into a context-rich, always-available assistant that dramatically improves call preparation and conversation quality.

• Pulls in CRM data, Gong call transcripts, emails, notes, and pricing — the more systems connected, the richer the context.
• Reduces manual prep time by an estimated 6+ minutes per interaction.
• When the assistant experienced its first-ever outage, Mark's team "lost their minds" — a clear sign of deep adoption and dependency.
• Key use cases: pre-call briefings, agenda creation, deck generation from QBRs, and risk-mitigation planning for at-risk customers.
• Improves conversation quality by enabling CSMs to arrive with tailored, customer-specific insights rather than generic updates.

Success Story 2: CS Outreach Agent

The Outreach Agent automates personalised email outreach at scale, enabling CSMs to book more customer calls without the time-consuming task of crafting individual messages. Launched just three months ago, it is already transforming how Mark's team starts each month.

• The agent drafts emails that pick up context from the last CRM conversation — including topic, tone, and sentiment — and writes in the CSM's personal voice.
• Recipients could not distinguish AI-generated emails from personally written ones, validating the quality of personalisation.
• CSMs begin each month by reviewing their customer lists, launching outreach agents, and quickly reviewing drafts — calls start flowing in and calendars fill up.
• Estimated time savings of up to 15 minutes per call when combined with features like AI-suggested contacts scraped from LinkedIn.
• Also used for targeted campaign execution, enabling the team to reach specific customer segments with contextually relevant messaging.

Success Story 3: Customer Support Agent

The third success story centres on HubSpot's Customer Support Agent — described by Mark as "a chatbot on serious steroids." This agent handles incoming support queries via email and web chat, and has already achieved remarkable deflection rates while improving response quality and speed.

• 65% of incoming support traffic is now handled by the agent, freeing human agents for higher-complexity work.
• No headcount reductions have resulted — instead, team members are redeployed to higher-impact activities.
• The agent detects customer frustration levels (tone, sentiment) and autonomously escalates to a human when appropriate, rather than waiting for the customer to request it.
• Supports seamless handoffs to sales teams with instant routing based on pre-configured rules, preventing lost revenue from slow follow-up.
• HubSpot uses the agent on its own support channels — "eating its own dog food" — to validate the product before recommending it to customers.

Key Takeaways and Recommended Next Steps

Mark closed with three takeaways and three actionable next steps, emphasising that getting started is simple and that the tools are already available within HubSpot's existing platform.

Three Takeaways

• AI is table stakes for CSM productivity — it is not a side project. Start small, build councils, and iterate.
• Breeze Assistant is essential for meeting new customer expectations. Every customer-facing team member — not just CSMs — should activate it immediately. It levels the playing field, especially for new hires ramping up.
• Agentic AI is how you scale repetitive jobs — agents should handle the mundane, admin-heavy tasks so humans can focus on connection and impact.

Three Next Steps

1. Identify use cases: Start with call prep, call wrap-up, or opening the assistant for every call in one team.
2. Start using Breeze Assistant this week: Experiment with prompts, find what works for your workflow.
3. Experiment with agents in your portal: Pre-built agents are available on Enterprise and Pro subscriptions with free credits. 28-day free trials are also available through your HubSpot rep.