Using AI to Build Actionable Audience Profiles for the New Museum

The Challenge

Capacity partnered with the New Museum to explore a practical question: could AI help turn the audience information they already had into a clearer, more actionable marketing strategy?

Rather than asking AI to simply generate personas, we built a research process that combined the museum’s existing audience knowledge with publicly available information. The goal was to identify the smallest number of audience profiles that would actually benefit from different marketing messages, then turn those insights into a tool the Museum could use in its day-to-day marketing.

The result: three distinct audience profiles grounded in evidence, plus a CustomGPT designed to help the Museum apply those insights to real messaging.

Starting With What the Museum Already Knew

The New Museum already had valuable audience research. Our first step was to understand what that research could tell us, and where there were gaps.

Capacity analyzed the Museum’s existing audience information for motivation, emotional drivers, identity, barriers, and engagement. We also assessed the strength of the available evidence and considered where bias in the underlying research could influence our conclusions.

Then we used AI to look for recurring psychographic patterns across those signals. For example, the analysis connected comments about contemporary art feeling difficult, uncertainty about “getting it right,” and positive reactions when an experience felt more accessible. Together, those signals pointed to confidence as a meaningful barrier for some audiences. The analysis also connected responses about inspiration, cultural pride, feminism, and representation, pointing to personal and cultural resonance as a meaningful motivation for engagement.

We required evidence for every pattern and flagged observations that weren’t adequately supported rather than treating them as audience truths.

This gave us a grounded picture of what the Museum already knew about its audiences before we expanded the research.

Using Deep Research to Expand the Picture

Internal research is especially valuable because it reflects an organization’s actual audiences. But it may not capture every perspective or potential visitor.

Capacity used AI-powered Deep Research to examine publicly available information about New Museum audiences. The research included reviews, cultural and lifestyle coverage, social and video content, and press and industry commentary. It surfaced perspectives ranging from highly engaged contemporary art audiences to casual visitors and people encountering the Museum as part of a broader cultural experience.

We then synthesized the external findings with the Museum’s internal research, looking for areas of alignment, meaningful differences, and insights that could affect marketing strategy.

The process used multiple AI platforms as part of Capacity’s research methodology, but that wasn’t a requirement for the Museum. This type of research can be conducted within a single paid AI platform.

Finding the Audience Differences That Matter

Audience segmentation can quickly produce more personas than a marketing team can reasonably use.

So we set a different standard: a separate audience profile needed to earn its place by changing how the Museum would communicate.

Our analysis considered differences in core motivations, emotional drivers, identity, barriers, and context. We gave the greatest weight to motivation and only separated audiences when doing so would materially change the message, tone, or emphasis.

The analysis ultimately supported three distinct audience profiles:

1. The audience that actively seeks intellectually and emotionally challenging contemporary art. 
2. Those motivated more by personal and cultural resonance.
3. People who approach contemporary art with curiosity but may need clearer signals that the experience is for them. For this audience, interest may already exist. Confidence can be the bigger barrier.

Those distinctions gave the Museum something more useful than demographic groupings. They provided a framework for thinking about why someone might engage and what could make an experience feel relevant to them.

Pressure-Testing the Profiles With Real Marketing

A useful audience profile should change the work. To test whether these distinctions were meaningful, Capacity applied all three profiles to the same New Museum exhibition. We explored how its positioning could change depending on the motivations, barriers, and interests of each audience.

The exercise demonstrated that the same exhibition could support substantially different messaging without changing the underlying offering.

It also helped the Museum identify applications beyond exhibition marketing. During the project, their team began exploring how the profiles could map to membership goals and inform acquisition and renewal messaging.

That was an important test of the work: the audience profiles could provide a shared strategic lens across different marketing needs rather than serving as personas that lived in a research document.

Keeping Human Expertise at the Center

AI allowed us to analyze a large amount of qualitative information, compare patterns, and explore potential audience distinctions efficiently—but it didn’t get the final say on who the New Museum’s audiences are. Human review was built into the process.

The Museum team reviewed the proposed profiles, challenged assumptions, and refined them using their firsthand knowledge of their audiences.

That combination matters. AI can help find patterns across complex information and make it easier to explore possibilities. Capacity brought the research methodology and marketing strategy needed to determine which insights were useful. The New Museum brought the institutional and audience knowledge needed to make sure the final profiles reflected their organization.

Turning Audience Insights Into a Working Tool

Once the audience profiles were defined and pressure-tested, Capacity built a CustomGPT to make them easier for the New Museum team to use in their everyday work.

We trained the tool on the audience insights, the Museum’s voice, and examples of its existing content. Instead of asking marketers to reference a research document every time they developed a message, the CustomGPT could apply the profiles directly to marketing tasks.

The tool was designed to help the team:

— Adapt messaging based on the motivations and barriers of each audience profile
— Adjust tone and calls to action for different audiences
— Develop audience-aligned content for priority marketing platforms
— Consider which audience profile was the strongest fit for a specific exhibition or event

The CustomGPT gave the Museum a practical way to carry the strategy into execution while keeping its team in control of the final work.

The Takeaway

AI has applications for arts and culture marketers that start well before content generation.

For the New Museum, AI-supported research helped turn existing audience information and broader public signals into three actionable profiles based on differences that could meaningfully affect marketing strategy. Capacity then built those insights into a CustomGPT that could help the Museum apply the profiles to real marketing decisions and content.

The result was both a clearer framework for understanding why different audiences might engage and a practical tool for putting that strategy to work, while keeping the Museum team’s knowledge and judgment at the center.

Want to explore where AI could make a meaningful difference in your marketing? Capacity’s AI Strategy services can help you identify practical applications grounded in your organization’s goals, data, and expertise.