AI Baked In - the Good Kind | Entry 97
HubSpot AI Context gives supported AI tools a shared understanding of the nonprofit's mission, audiences, programs, terminology, priorities, and ways of working. Instead of repeating the same background in every prompt, teams can create an approved organizational reference layer that helps Breeze produce more relevant content, recommendations, and assistance.
AI needs the information that makes the organization distinct before it can assist in a way that reflects the nonprofit's actual work.
AI Context provides organizational background that supported AI experiences can use throughout HubSpot. It helps AI understand what the nonprofit does, why the work matters, who it serves, who supports the mission, how teams operate, which goals matter, and which language is appropriate.
It is an organizational reference layer, not a replacement for precise prompts, current CRM data, or human expertise.
Source information should be accurate, current, approved, and safe to use. AI Context supports human decision-making, but it does not replace fact-checking, subject-matter review, constituent consent, or organizational accountability.
A nonprofit centered on community-led solutions may want generated content to emphasize participation and partnership instead of portraying the organization as the sole source of change.
Useful program context can include names, approved descriptions, eligibility requirements, service locations, application steps, schedules, participation requirements, verified impact claims, referral details, contact channels, and frequently asked questions.
Review this information whenever programs, eligibility rules, schedules, or service areas change.
Each audience has different goals, questions, motivations, communication needs, and relationships with the organization.
Responsible audience context may cover common motivations, programs of interest, preferred channels, gift types, frequent questions, concerns, appropriate calls to action, and stewardship approaches.
The objective is to help AI understand a type of audience, not to place unnecessary personal details about individual donors into organizational context.
Context should clarify respectful terminology, relevant services, approved eligibility language, appropriate channels, information that must not be used for marketing, topics requiring human support, and situations involving consent or guardians.
One organization may use “program participant,” another may use “client,” and another may avoid both. Approved language reduces editing and improves consistency.
Provide approved information such as people served, services delivered, communities reached, programs completed, funds distributed, volunteer hours, documented outcomes, annual-report statistics, research findings, and evaluation results.
| Information type | How AI should treat it |
|---|---|
| Verified outcomes | Use only with the approved scope, date, and source |
| Organizational goals | Describe as intended future results |
| Estimates | Label them clearly as estimates |
| Aspirational statements | Do not present them as completed outcomes |
| Historical results | Preserve the correct reporting period |
| Current results | Confirm they remain accurate before publication |
AI should never invent a statistic or transform an aspiration into a completed result.
AI can then understand that major-donor inquiries go to development, volunteer questions go to the volunteer coordinator, and eligibility questions need program-team review.
Nonprofits may connect HubSpot with accounting, donation, event, volunteer, wealth-screening, direct-mail, learning, payment, document, grant, and communication systems.
| Example system | Possible source of truth |
|---|---|
| HubSpot | Constituent engagement and relationship activity |
| Accounting platform | Recognized revenue and financial records |
| Volunteer platform | Shift schedules and service activity |
| Event system | Ticketing and attendance scanning |
| Connected applications | Selected synchronized data available in HubSpot |
This prevents AI from assuming that every organizational process occurs directly in HubSpot.
AI may benefit from knowing about similar organizations, partners, coalitions, industry associations, community resources, public agencies, policy organizations, influential voices, and sector trends.
This should support accurate positioning, not misleading comparisons or unsupported claims about other organizations.
Every source needs an owner and a maintenance process. Outdated knowledge can generate confident but incorrect answers.
| HubSpot layer | Question it helps answer |
|---|---|
| AI Context | What does the organization do, and what matters? |
| Brand Voice | How should the organization communicate? |
| CRM data | What is known about this constituent or relationship? |
| Knowledge bases | Which answers and procedures are approved? |
| Workflows | What should happen under these conditions? |
| Agents | Which activities may AI perform? |
| Permissions | Who and what may access the information? |
AI Context explains who the nonprofit is, what it does, who it serves, how it operates, and which information matters. Brand Voice explains how the organization should sound, which tone and writing characteristics to use, and how content should feel.
Context without Brand Voice may produce informed but inconsistent writing. Brand Voice without context may produce polished content that lacks relevant substance.
These layers can work together without exposing unnecessary personal information.
With complete and current context, staff may be able to ask which team handles sponsorship agreements, where the volunteer-onboarding guide lives, which description of a program is approved, which system owns payment information, who approves campaign content, or what follows a major-gift inquiry.
Context can reduce generic language, but it cannot prevent inaccurate claims or inappropriate recommendations by itself. Externally facing content still needs review.
Without ownership, context can become a collection of conflicting, duplicated, or outdated information.
Do not add sensitive donor, beneficiary, volunteer, employee, financial, legal, medical, or personally identifiable information unless the organization has confirmed that the use is appropriate and permitted.
More information does not always produce better results. Prioritize accurate, approved, current, clearly organized information that is relevant to defined AI use cases and has a named owner.
Remove obsolete policies, outdated program descriptions, expired events, duplicated resources, and unapproved drafts from the sources AI relies upon.
Effective context should reduce repetitive explanation and revision while improving the usefulness of supported AI tools.
Availability can vary by subscription, product, language, permissions, region, beta participation, and rollout.
With approved knowledge, clear ownership, thoughtful permissions, and ongoing maintenance, AI Context can help HubSpot understand the nonprofit's mission, audiences, programs, language, systems, and priorities without requiring teams to explain the same foundation every time.