After seven years as a HubSpot partner, this was my first time attending HubSpot’s annual conference in Boston.
Technically, of course, it was everyone’s first UNBOUND.
After fifteen years as INBOUND, HubSpot changed one letter. But there was a lot of meaning behind that one-letter change.
INBOUND was born around a particular way of growing through marketing. UNBOUND reflects how much that world has expanded: marketing, sales, service, operations, AI, data and the entire customer journey are increasingly connected. As Yamini Rangan explained during Spotlight, the way we grow has fundamentally changed, so the event needed to evolve with it.
I think that was more than clever rebranding. It became a perfect lens for the whole week.
I went expecting to come back with product updates, ideas for customers and plenty of notes. I did come back with all of those things. Through our UNBOUND 2026 Intelligence project, we have been unpacking sessions across AI, GTM, marketing, sales, service, data, customer experience, leadership and the future of work.
But once I got home and started going back through everything, the individual sessions began to blur together a little, and a much clearer set of themes emerged.
AI is making it easier to research, create, automate and build. So the interesting question is becoming less about access to AI and more about what happens when everyone has it.
Where does the advantage come from then?
For me, UNBOUND kept coming back to the same answer from different angles: clearer outcomes, better context, more selective use of AI, stronger human judgment, greater trust and a much more deliberate view of where the business is heading.
In many ways, that feels aligned with the change from INBOUND to UNBOUND itself; moving beyond one methodology, one channel, or one function and thinking about growth as a more connected system.
In a nutshell, these were my main takeaways from UNBOUND 2026: start with business outcomes, focus AI on fewer high-value use cases, treat context as infrastructure, protect the value of human judgment, build for trust, follow the changing discovery journey, design agents into real operating processes and make enough space for strategic foresight so that the speed of AI does not start setting your strategy.
One of the numbers from HubSpot Spotlight that stayed with me was this:
90% of companies are using AI, but only 6% are seeing transformational results.
HubSpot's research covered 6,000 customers and prospects across multiple industries and around 50 AI use cases. The 6% seeing transformational results were also four times more likely to hit their revenue targets and three times more likely to hit their efficiency targets.
That was probably the point where the conversation about AI maturity shifted for me.
We have spent the last few years talking about adoption. Which tools are companies using? How many pilots are running? How many agents have they built?
But if almost everyone is already using AI, adoption itself tells us very little.
During Spotlight, Yamini Rangan described this as a kind of "maxing" phase: more tools, more pilots, more tokens, more activity.
The next phase is what HubSpot is calling the Outcomes Era.
Yamini has since expanded on it in The Outcomes Era is here, and I think the core idea is useful well beyond HubSpot: start with the business outcome and work backwards.
What are we actually trying to improve?
Pipeline generation? Conversion? Seller productivity? Retention? Customer satisfaction? Response times? Forecast accuracy?
Once the outcome is clear, we can work backwards into the process, data, context and technology required to change it.
That sounds obvious when written down. In practice, it is a useful test because it exposes how easy it is to start with the technology instead.
This was probably the finding from Spotlight that made me think most about our own AI roadmap.
The companies seeing transformational results were not doing more AI. They were doing fewer use cases. That is almost the opposite of the pressure most businesses are feeling right now.
Every week there is a new model, agent, connector, feature or workflow to explore. You see somebody demonstrate an impressive use case and immediately start wondering whether your business should be doing it too.
It is very easy to accumulate experiments.
And after a while, the number of AI initiatives starts looking like a measure of progress even when nobody can clearly explain what changed because of them.
The lesson I took from this was to be more selective. Which few use cases could genuinely change something important in the business? Then give those use cases enough attention, data, ownership and context to work properly.
I came away thinking that good AI strategy may increasingly be about restraint. Knowing what not to pursue could become as important as knowing what to build.
The conversation about context immediately took me back to the HubSpot portals we work in every day.
Imagine two teams cannot agree on what an MQL actually means. Giving an AI agent access to the MQL property does not suddenly make that process intelligent. The agent has the data point. It does not necessarily understand the business meaning behind it. That distinction became one of the most important ideas for me from Spotlight.
Yamini describes context across three areas: your business, your customers and your team.
Business context is your products, goals, positioning and brand. Customer context is the relationship history, ICP, conversations, needs, sentiment and commercial information surrounding that customer. Team context includes roles, permissions, workflows, handoffs, approvals and the way work actually gets done.
HubSpot's research also showed a dramatic difference in results when AI was paired with good context. Its analysis found improvements of 264% in MQLs generated, 197% in deals won and 200% in customer meetings booked. With poor context, those same measures moved in the wrong direction.
This matters because a lot of what we already think of as good RevOps is really context.
Your ICP is context.
Your lifecycle definitions are context.
Your lead source framework, qualification criteria and sales methodology are context.
Customer history is context.
Governance, permissions and escalation rules are context.
So is the knowledge currently living in somebody's head because nobody has documented how a process actually works. That is why I came away from UNBOUND more convinced that better AI makes CRM foundations more important. If the underlying process is unclear, the AI inherits that ambiguity.
For us, AI readiness increasingly looks a lot like good RevOps. Clean data, clear definitions, documented processes, appropriate governance and enough context for both people and AI to understand how the business operates. That is also why HubSpot's direction around Growth Context and the self-updating CRM interests me more than another list of AI features.
We went much deeper into that in our HubSpot Spotlight: The Outcomes Era breakdown.
One thing I liked about UNBOUND was that this idea did not come from a single AI session.
It kept appearing in completely different conversations.
Dharmesh Shah talked about the age of the builder. Grant Lee explored what it takes for ideas to actually land. Sales sessions challenged the assumption that more automation automatically creates a better selling experience. Data leaders talked about the infrastructure needed to make AI useful rather than simply impressive.
I kept connecting those conversations back to the same thing. AI is lowering the effort required to produce.
Research can happen much faster. Content can be generated almost instantly. Workflows can be prototyped without traditional development resources. Agents can take on work that previously required considerable manual effort.
That is a huge shift, but when production becomes easier, producing something is less of an advantage by itself. If everyone can generate a first draft, the first draft is worth less. If every company can publish more, volume is less interesting. If everyone has access to increasingly capable models, access to the model cannot be the differentiator forever. This is where the human part becomes more important, not less.
Judgment. Perspective. Taste. Empathy. Creativity. Experience.
The things that help us decide whether something is worth doing in the first place, whether it is right for this particular customer, and whether the output is actually good.
My takeaway was not that humans need to compete with AI at execution.
It was that we need to become clearer about where human thinking creates the most value, then use AI to multiply it.
This connected with another idea that surfaced repeatedly across the week.
AI has made information and content abundant.
Attention was already difficult to earn. Now buyers also have to navigate more generated content, more automated outreach and more perfectly polished messages than ever.
That makes trust much more valuable.
Maha Abouelenein talked about storytelling and trust in a skeptical world. Shana Sumers challenged the usual idea of networking and focused instead on relationship capital. GTM sessions kept coming back to understanding the buyer deeply enough that they can recognise themselves in what you are saying.
Different topics, but the underlying principle was similar. People want to know who the information is coming from.
Does this person actually understand the problem? Have they experienced it?
Do they have a point of view? Would I trust their advice?
Subject-matter experts, communities, customer evidence, personal brands, and direct relationships will matter more. People still follow people. AI can help scale communication, but credibility still has to be earned.
For marketers, that means efficiency cannot be the only measure. We also need to ask whether what we are creating is useful enough to remember and credible enough to trust.
Discovery was another theme that appeared in several different sessions rather than one big keynote moment.
People still use Google, of course. But they are also researching through ChatGPT, AI search experiences, Reddit, YouTube, LinkedIn, communities and other sources.
At UNBOUND, HubSpot's AI Search Lab shared experiments around AEO. The Own Your Audience session made the case for publishing directly and building relationships with audiences, rather than relying entirely on rented distribution. GTM sessions reinforced the importance of positioning and understanding the person behind the search.
When you connect those ideas, marketing starts to look broader than traditional search optimisation.
Ranking for a keyword still matters, but so does whether your expertise appears in the places buyers use to answer their questions before they ever reach your website. That means first-party content, genuine subject-matter expertise, customer evidence, third-party authority, social content, communities, original research and clear positioning all become part of the discovery strategy.
The goal is to be discoverable, credible and useful wherever the buyer is doing the research, and SEO remains part of that; the surface area around it is getting much bigger.
There were agents everywhere at UNBOUND. The more sessions I attended, though, the less interested I became in how many agents a company might eventually use.
I became much more interested in what happens when an agent becomes part of an actual business process.
Once an agent can research, analyse information, communicate, recommend an action, update a CRM or trigger the next step in a workflow, the conversation changes.
Now we need to know what information it can access.
What is it allowed to change?
Which decisions can it make?
Where does a person need to approve something?
Who owns the process if something goes wrong?
How do permissions work?
How do we know whether the agent is improving the business outcome it was supposed to improve?
That is where sessions about HubSpot's open platform vision, AI Work Management, data strategy and agentic workflows started connecting for me. Individually, they were product and technical conversations. Together, they point towards agents becoming participants in business processes rather than isolated AI tools sitting alongside them, and that brings us straight back to architecture, data, permissions, governance and process design.
We have captured those individual sessions across UNBOUND 2026 Intelligence, and we will cover the specific HubSpot agent and platform releases separately.
One of the sessions I have thought about most since coming home actually happened before UNBOUND officially started. At Partner Day, Amy Webb delivered a keynote on strategic foresight.
By that point, I had already spent weeks thinking about everything changing in AI. Her keynote made me realise that constantly trying to keep up can become its own trap.
Every week brings another announcement, capability, prediction or new way of working. The natural response is to move faster. But are we moving fast because we know where we want to go, or because everything around us is moving fast?
Strategic foresight asks us to look beyond the immediate announcement and think about where customer behaviour is moving, where value might be created, which capabilities we may need and which signals would tell us that our assumptions are changing.
And this is where Amy's keynote connected, for me, with Yamini's Outcomes Era.
Strategic foresight asks where we are going. The Outcomes Era asks what result we are trying to create.
Then we can work backwards and ask what we should start building or changing now.
I like that combination because it avoids two extremes: chasing everything that changes and creating a five-year strategy so rigid that we cannot respond when the evidence changes.
We still need to execute well today.
We also need to deliberately create enough space to think beyond today.
Some of my favourite parts of UNBOUND had very little to do with software.
We heard from an astronaut. A behavioural change expert. Entrepreneurs. Creators. People thinking about leadership, communication, relationships, storytelling and performance from completely different perspectives.
I loved that HubSpot did this.
When everybody in an industry reads the same playbooks, follows the same experts, attends the same conferences and increasingly uses the same AI systems, ideas from outside that bubble become incredibly valuable.
Sometimes the thing that changes how you think about marketing comes from someone who is not a marketer. Sometimes a lesson about leadership comes from someone whose working environment could not be further from yours.
That was one reason the event felt broader than a product conference to me, and perhaps that is another expression of what the name change was trying to capture. After fifteen years of INBOUND, UNBOUND feels less like abandoning what came before and more like widening the frame. HubSpot describes it as the next chapter of the same community, but built around the much broader reality of how companies grow today.
AI is making information abundant.
Perspective is still much harder to manufacture.
We came back from Boston with far more ideas than we could realistically implement. I do not think the objective should be to implement all of them. In fact, that would go against one of the clearest lessons of the week.
For me, the more useful exercise is to bring the learning back into the business and ask better questions:
Those are the questions I am bringing back into Cat Media, into our work with customers and into the conversations we continue to have with the Dublin HubSpot User Group.
They feel much more useful than asking which AI tool we should add next.
This article is the synthesis. We captured much more while we were in Boston.
Our UNBOUND 2026 Intelligence experience brings together the sessions we covered across AI, GTM, marketing, sales, service, customer experience, leadership, data and the future of work, including full session breakdowns, short audio digests and stories from the experience on the ground.
Yamini Rangan has also expanded on the thinking since UNBOUND in The Outcomes Era is here.
And if you want to explore the product announcements directly from HubSpot, the official HubSpot Spotlight brings together the latest product launches and updates announced at UNBOUND.