This is part three of a three-part series on how HubSpot has transformed itself with AI. The first part explains how we build with artificial intelligence. Part two illustrates how we grow with Agent-first GTM.
Building the right design platform and rebuilding the go-to-market process is meaningless if the organization managing it isn’t ready. This is the part that most transformation manuals skip. It’s also the part that determines whether something sticks.
We didn’t skip it; we doubled. As a result, 94% of HubSpotters use AI weekly, employees have created over 3,900 AI agents, and our talent profile looks fundamentally different than it did three years ago.
This is our playbook for the HubSpot organizational transformation that made everything else possible.
Phase 1: Develop AI mastery (2023-2025)
The first phase is about fluidity across the entire organization and must start with commitment from the top. Leaders need to model behavior, share their experiments, and create conditions for everyone else to follow, not obligations.
To get to this we ran three representations, each of which is repeatable for any organization:
Provide the teaool. Each HubSpotter received enterprise licenses for a core set of AI tools. A central AI strategy team manages supplier relationships, defines security standards, and streamlines the adoption of new tools, eliminating the procurement and security bottlenecks that slow transformation in most companies. The fluidity of artificial intelligence cannot be a competitive advantage reserved for certain teams. It must be a basic expectation for all teams.
Change your mentality. This included fostering a culture of experimentation, where employees feel empowered to try and embrace new ways of working. We’ve updated our company values to encourage this perspective, adding “be bold, learn fast” as a core value. Employees share use cases and experiments in dedicated chat channels. Leaders participate alongside their teams, often receiving reverse mentorship from people further along in their experimentation, and executives share their learnings in weekly updates. We also changed the speed of our organizational clock from annual planning cycles to six-week sprints to keep up with technology.
To track our progress, we also set a clear, company-wide usage goal: 80% active weekly AI usage by the end of 2025. Then we tracked it openly – by team, by tool, by use case – and made the data visible to everyone. Transparency spurred accountability in both directions: Teams that were behind received a clear signal, and teams that were ahead became role models for others.
We want to be clear about why we tracked usage rather than outcomes at this stage. Phase 1 was about developing the fluency of the AI. You can’t measure improved outcomes with tools that people don’t yet use. Usage was an important indicator, not destination. This wasn’t tokenmaxxing; it was a necessary step on the way to maximization of results in Phase 2.
Develop skillsT. We carved out protected time for learning. This included hackathons and 20 enterprise-wide AI learning days in 2025. AI was integrated into onboarding from day one and into ongoing manager development. The goal was simple: move the question from “should I use AI for this?” to “how can I use AI better?”
The result of Phase 1 was a new talent profile. By the end of this phase, we had an organization that was becoming fluent with AI, with 94% of HubSpotters using AI weekly, with over 3,900 AI agents created by employees to improve their work.
Phase 2: Team-wide transformation (2025–present)
When employees use AI in different ways for different use cases, it results in individual productivity but not business results. To achieve team-wide transformation, you need clear priorities with real accountability behind them.
To start, we plotted the teams against two dimensions:
- AI maturity: How did they adopt the tools? Are they seeing measurable results?
- AI readiness: What is the potential of the team’s work for automation? Is there business risk? Are the data infrastructure and AI capabilities there to support?
This analysis produced three categories for us: Rhythm setteror teams that were already moving quickly. We don’t want to slow down these teams; we want to support them. He almost winsor teams that have clear automation opportunities but have not acted. The bottleneck for these is almost always leadership attention, not tools. And finally, Big bets. These are the teams with the highest potential but the most dependencies. They need dedicated investments in data, systems and change management.
Here’s where our teams fell, each requiring a different playbook:
Pace Setter: Engineering, Support and Marketing had already seen major improvements in productivity and efficiency through proven use cases of artificial intelligence, leadership sponsorship and measurement. They needed minimal support and continued their momentum through investments in AI fluency.
Marketing is the clearest example. The team reinvented workflows at every level: AI-powered email personalization led to an 82% improvement in email conversions, an AI-powered chatbot now handles over 82% of website inquiries, and generated over 10,000 sales meetings per quarter by Q4 2025. A video ad production test produced AI-generated spots at $300-$ 3,000 compared to $300,000-$500,000 with traditional production and AI-assisted blog production. reduce writing hours per article by 60%.
Close victories: recruitment and operations had clear automation opportunities that could be unlocked with the right tools. The key lever was leadership focus: “gemba walks,” getting to work with teams to identify exactly where AI could replace or augment specific tasks, and driving adoption in the field rather than remotely.
An example of this is talent acquisition. By incorporating AI directly into the hiring funnel, we saw a 10-day reduction in time to hire and a 30% reduction in application review time. We fully automated 80% of interview scheduling tasks, resulting in a 90% increase in scheduling volume without additional staff. The percentage of hires coming from legacy candidates grew from 8% to 18% in the first 90 days, a direct result of AI resurfacing talent that would otherwise have remained invisible.
Big bets: Sales, CCustomer success and product has the highest potential but needed significant investment in data, systems and change management. These teams received dedicated AI pods, cross-functional teams of functional experts, data scientists, and operations engineers focused on reimagining specific workflows through rapid experimentation and iteration.
The deeper lesson of Phase 2 is that not all teams need the same support. Maturity and readiness analysis is what tells you where to push, where to support, and where to invest. Without it, you end up applying the same approach everywhere and wondering why only some work.
Phase 3: Institutional transformation (2026 and beyond)
We are at the beginning of Phase 3. But the direction is clear and it will be the most important phase of all.
Phases 1 and 2 solved for individual and team productivity. Phase 3 is about creating institutional AI. The distinction matters. Making each employee 10x more efficient does not make a company 10x more productive, unless the institution itself is redesigned around new AI capabilities.
The foundation of Phase 3 is the institutional context. It means giving everyone access to the right tools, data and information, and codifying business processes into agents that can act on them at scale.
The difference becomes visible in the way the work is done day to day. When an engineer needs context about a code base, he doesn’t ask a colleague; they ask HubSpot’s internal coding agent. When a sales manager wants to understand why a deal stalled, they don’t write a report; they ask our native Guided Sales Assistant. When a new hire needs to understand how HubSpot makes decisions, they don’t wait until onboarding; they ask our in-house AI tool. This is what institutional AI looks like in practice: the collective context of the organization, available to everyone, when they need it.
Moving to this stage also requires addressing issues that previous stages do not address. When agents own the steps of an end-to-end workflow, governance matters more. Who can see what? Which decisions require human approval? How do you detect bad results before they get worse? We had to deliberately answer these questions, establishing clear authorizations, audit trails and escalation paths so that the speed of agents does not exceed our ability to supervise them.
We are still on this journey. But we understand what is at stake. Companies that build institutional AI are the ones that will have an advantage. But to do that, don’t start with artificial intelligence. Start with work. Find the workflow slow, expensive, or fragile. Find the most ready team. Run the experiment, measure it honestly, and then commit to what the data shows.
AI transformation starts with a solid foundation
The same principle runs through everything in this series: the tools are just the starting point. Building the foundation – technically, structurally and culturally – is what allows you to grow.
In engineering, that foundation is a platform. In the go-to-market, it’s a flywheel. In the way you operate, it is the organization itself. Companies that understand this will not only use AI better, they will grow better.
