Artificial intelligence was supposed to make your workload lighter. For most marketing and sales teams, the opposite has happened: you’re launching more campaigns across more channels than two years ago and doing your best to keep up with the latest AI tools between sales calls. The volume of work has grown exponentially, yet for many companies, their outcomes haven’t.
Recent HubSpot research cited at UNBOUND 2026 reports that 90 percent of companies reported using AI, yet only 6 percent have seen transformational results. While AI adoption has doubled, many companies cannot produce the metrics that show ROI. But the ones that can are four-times more likely to hit their revenue goals and three-times more likely to hit their efficiency targets. Which begs the question: what do the companies transforming with AI have that others lack? And what are they doing differently?
We’ve observed that the biggest obstacle to achieving great AI outcomes has little to do with your team’s capability or the AI tools at your disposal and everything to do with bad CRM data.
How Bad CRM Data Affects Your AI Outcomes
The whole point of having robust, high-quality CRM data is to help you effectively speak to a cast of thousands in an individualized, one-to-one manner. When we say “CRM data” we’re talking about everything you collect about your customers, from contact details and behavioral signals to deal records and support tickets. Bad CRM data, AKA data that’s fragmented, outdated, unreliable, and/or duplicated, has long been the enemy of marketing and sales teams, and poses real problems when integrating AI into workflows.
Common data hygiene problems we see include:
- Data duplication or “dirty data”
- Data siloed outside of the CRM
- Contacts with incorrect lifecycle stages
- Missing company associations
- Inconsistent property values
HubSpot’s CEO Yamini Rangan declared at UNBOUND 2026 that “bad context is worse than having no AI at all.” We couldn’t agree more. By training AI on bad data, you amplify it. When a contact record or AI output is wrong, a human normally notices and can intervene, whereas an AI agent often doesn’t do either. An error someone could have caught can be repeated across workflows and across hundreds, if not thousands of customer interactions, which damages brand trust.
Your CRM is the foundation of your GTM. Yet we see much of the richest data residing outside the CRM, in call transcripts, email and Slack threads, shared drives, and meeting notes. A CRM lacking this unstructured data leaves your AI tools working from an incomplete picture of your business, and more importantly, your customer. If you don’t tell the CRM what it should know, it won’t work with the knowledge you and your team hold.
As HubSpot expands its AI agents across marketing, sales, and service, the need for complete, reliable, robust CRM data becomes much more pressing. Treat your data as a distinct business capability.
“Bad context is worse than having no AI at all.”
— Yamini Rangan, CEO at HubSpot
The Context AI Needs From Your Business
HubSpot’s agents are built on your CRM and the magic lies within the context of your business. Building good context into HubSpot will ensure your human and AI teammates work well with the platform and create great outcomes. By leveraging context, AI starts acting like an expert of your business instead of a generic robot with a superficial understanding of your business. You have to treat AI like a new employee and teach it everything it needs to know to do its job well.
That begins with providing access to deep, specific, current data you and AI can trust, grouped into three categories:
- Business context. The description of what your business does and what makes it unique, for example, your brand story, its tone of voice, and product positioning.
- Customer context. The range of interactions happening across channels, the personas you target, and their buying signals.
- Team context. The people on your team, their roles and goals, and ways of working together.
We call it “growth context”: the shared, dynamic understanding of your entire business.
Once each layer of context is complete inside HubSpot, what you can achieve with the tools becomes infinitely better. Prospecting Agent researches high-value accounts and drafts outreach that reflects actual buying signals. Customer Agent can answer service inquiries by running off your approved knowledge base and existing ticket history. Data Agent can also pull insights from your sources and the web to populate records with smart properties. The possibilities for transformation are endless.
What Teams Transforming with AI Do Differently
Every GTM team has access to the same tools. AI’s ubiquity lowered the barrier to building and creating just about anything. However, the teams successfully transforming how they go to market and scale with AI are making purposeful choices, rather than executing AI projects for the sake of it. They are focused on building a solid data foundation of rich, reliable context that drives the outcomes they are targeting.
What’s more, transformational teams solve business problems and achieve specific metrics, instead of worrying about optics. Interestingly, they have their sights set on the goals they’ve chased for the last decade: building demand, winning deals, and delighting customers. That’s nothing special. But findings from that aforementioned HubSpot research showed the most successful companies focus on a handful of use cases and run them well. Read: they are selective.
Their highest value use cases rely on rich data and context, such as analyzing campaign performance, flagging at-risk deals, or managing customer offboarding. Meanwhile most ask AI to create content or send sales and service emails. These popular tasks need little growth context in order to execute.
The difference between transformational outcomes and a failed project boils down to leveraging knowledge about your business. This unique knowledge, which competitors cannot copy, builds AI’s understanding. Wield your growth context, and it becomes your distinct advantage.
Where to Start with Your CRM Data
Before you add another tool to your AI arsenal, answer three questions. First, which outcomes do I want to improve? Do I have the data and context to achieve those outcomes? And do I have people who can think critically and build with AI? Most marketing and sales teams find the second question needs attention. Auditing your HubSpot CRM to understand where your instance excels and its limitations can help spot gaps and inconsistencies in growth context, and show how much of your data you can trust.
Ask us for a comprehensive assessment of your HubSpot portal and we’ll uncover ways to unlock its full potential and maximize the outcomes you want from AI. Sound good? Talk to us.
