HubSpot's AI maturity guide caps your score at one level above your weakest area. Score each workflow, not the whole team

HubSpot published a way to score AI maturity, and the useful part is a rule that stops teams from grading themselves too high.

What happened

HubSpot's new guide on AI maturity draws on its 2026 State of AI in Marketing research of 1,754 marketers. It finds 98% of marketing teams use AI in some way, but only 10% report transformational results. Marketing leads other functions: 42% of marketers report advanced AI use, against 27% in sales and 23% in service. HubSpot's wider study of over 6,000 professionals found 90% say their company uses AI and 6% rate it transformative. HubSpot notes these are self-reported figures and says to treat them as directional.

The guide describes four stages: Early, Emerging, Established and Transformational. It also offers an editorial scoring framework, which it says is not an official HubSpot metric. You score six dimensions from 1 to 4: use cases, data and CRM context, workflow integration, governance, measurement, and strategy and ownership. Two rules stand out. Your effective stage cannot sit more than one level above your lowest dimension. And if you cannot produce evidence for a dimension, such as a run log or a written policy, you score it one level lower.

My take

The cap rule is the best part. I regularly see teams with ten AI tools and no idea which ones touch the CRM. On paper they look advanced. Under this rule, their governance score drags them back to where they really are.

The guide also says to score the workflow first, and that is how I would use it. For each automation:

1. Can it read clean, permissioned CRM data, or is it working from exports?
2. Is it inside the real process, or running beside it?
3. Is there a run log someone actually checks?
4. Can you show a before and after number?

The lowest answer tells you what to build next. Usually it is data cleanup or measurement, not another AI feature.