In Mass Culture’s most recent DNA cohort, festivals made up nearly one third of participating organizations. From the outset, we were struck by the enthusiasm festivals showed for engaging more deeply with their data, and we wanted to keep a close eye on where that curiosity might lead.
As the cohort journeyed through the work with their data coaches, many participants were asking themselves similar questions of their data and seeking similar kinds of insight. Yet the paths they took to get there varied widely. Festivals in particular were developing distinct strategies from one another to better understand many of the same underlying questions and curiosities.
Through regular check-ins and reflections in the Data Fellows’ learning logs, recurring themes surfaced: questions about investing in CRMs, better understanding audiences, and building sustainable systems for collecting, sharing, and learning from data over time.
To dive more deeply into some of these recurring themes, Robin Sokoloski, who manages the DNA project, brought together staff from SummerWorks and the PuSh International Performing Arts Festival for a shared data exchange. What emerged was a candid and generative discussion about the challenges festivals face and the possibilities that open up when festivals carve out the capacity needed to work with their data.
Many of the ideas explored in that conversation extend well beyond the festivals and resonate across the broader arts community. So, we wanted to share some of that dialogue with you here.
The Conundrum (We All Face)
Festivals operate within an intensely cyclical rhythm. While the work continues year-round, much of it builds toward a concentrated annual moment: the festival itself. Many arts organizations work in a similar cyclical fashion, but festivals experience this pattern in a particularly heightened way. Their structures, staffing, timelines, and decision-making processes are all shaped by the rapid build-up, delivery, and aftermath of a single large-scale event.
This cyclical way of working also has significant implications for staffing and organizational memory. High turnover among seasonal and contract staff can make it difficult to maintain the long-term systems and continuity to allow data to do what it does best. At Mass Culture, we often say that impact cannot be understood in the immediate. It reveals itself over time. But for that to happen, organizations need structures that allow knowledge, processes, and data to carry forward from one festival cycle to the next.
The truth is that all arts organizations struggle with the capacity required to maintain healthy and sustainable data systems. By sharing the insights from this festival data exchange as a kind of living case study, we hope to surface practical ways organizations can navigate some of these persistent and unavoidable challenges together.
Start With the Story, Not the Software
A core lesson from both Gaëlle Ramboanasolo’s work with SummerWorks and Meena Das’s work with PuSh Festival is one that sits at the heart of Mass Culture’s own approach to data: the most useful data practices begin with a meaningful question. SummerWorks did not enter the process by asking which tool or platform would solve its data challenges. Instead, its questions emerged from values-led management practice: How can we better understand and communicate the value of accessibility measures, community engagement, artist relationships, and organizational continuity? How can we tell a stronger story about work that is often relational, time-consuming, and difficult to capture through numbers alone? From that starting point, SummerWorks identified a practical goal for the coaching process: to create user-friendly, nimble systems for data capture, analysis, and sharing across audience and community engagement, artist contracting, and succession planning. The lesson is transferable: before investing in systems, arts organizations need to know what they are trying to understand, who the insight is for, and how that knowledge will support future decisions.
PuSh Festival entered the DNA cohort from a different starting point. Unlike SummerWorks, which was focused on connecting information spread across multiple platforms, PuSh had already invested in a dedicated CRM and was in the process of consolidating organizational data. The question was not whether they needed a system, but how to make that system more useful.
Working with data coach Meena Das, the team focused on strengthening the practices surrounding data collection, audience engagement, and organizational learning. A key area of work involved designing more strategic and consistent ways of reaching patrons through regular surveys, enabling the festival to gather richer audience insights while building stronger relationships with its community. The team also began developing new approaches to segmenting patron records, allowing for more targeted outreach and more meaningful post-festival data collection. At the same time, they explored how information generated during the festival—including attendance figures, audience feedback, staff observations, and event-level learnings—could be captured in real time rather than reconstructed months later from memory. Together, these efforts helped shift data collection from a retrospective reporting exercise toward an ongoing practice of learning and engagement.
PuSh’s experience highlights an important lesson for arts organizations: investing in a CRM is not the same as establishing a data practice. Technology can provide a foundation, but meaningful insight depends on the processes, habits, and organizational culture built around it. Even with a sophisticated system in place, organizations must still determine what information matters, who is responsible for collecting it, and how that knowledge will be shared and carried forward from one cycle of activity to the next.
Many organizations begin their data work by asking, “which platform should we buy?” A more useful starting point is, “what do we need to understand, explain, or decide?” Once that question is clear, the tools become easier to choose and the data becomes easier to interpret.
TRANSFERABLE TIP: Before investing in a new CRM, dashboard, survey tool, or AI workflow, write down the three stories your organization most needs to tell with confidence. For example: Who are we reaching? Who is returning? Which communities are engaging with us in more than one way? What evidence helps us explain the value of work that is not easily captured by ticket sales?
Build a data map before building a dashboard
SummerWorks’ dashboard work surfaced a reality many arts organizations will recognize: important information often lives in many places at once. Submissions may be in Airtable. Ticket purchases may be in a ticketing system. Donations may sit in CanadaHelps or another fundraising platform. Mailing list data may live in Mailchimp. Artist, industry, partner, audience, and volunteer information may each be tracked in different spreadsheets, surveys, or project files.
Rather than pretending one system could instantly solve that fragmentation, the SummerWorks team began by identifying where each kind of data lived and what needed to happen to make it usable. They built a cleaned data structure that could feed into Looker Studio, a free Google dashboarding platform. Once the cleaned spreadsheet was connected, the dashboard could refresh as new data was added, allowing the organization to see patterns across submissions, ticket buyers, donors, artists, and other communities over time.
TRANSFERABLE TIP: If much of your organization's data already lives within Google Workspace, it may be worth exploring Google Looker Studio before investing in a CRM. As SummerWorks demonstrated, powerful dashboards and reporting tools can be built by connecting existing data sources and creating clear processes for maintaining them. For those looking to get started, Arts Etobicoke's Digging Into Data Self-Guided Learning Program includes an excellent step-by-step tutorial on using Google Looker Studio. And if you need inspiration, SummerWorks' Kass Prus has become something of a Looker Studio aficionado (from our perspective) through this process. https://diggingintodata.artsetobicoke.com/analysis-and-visualization/
The transferable practice here is simple but powerful: make a data map. List each data source, what it contains, how it is collected, how often it is updated, what format it exports in, and where the cleaned version should live. This map becomes a bridge between people, platforms, and annual cycles.
As a side note, another organization in the DNA cohort, Wonder’neath Art Society, worked with Gaëlle to create an impressive visual data map of their organization. The map helps illustrate how information flows across programs, relationships, and decision-making processes. If you’re curious to see what a visual data map can look like in practice, let us know and we’d be happy to connect you with them.
Design for staff turnover and handoff
Both festivals spoke directly to the realities of seasonal staffing, role transitions, and limited capacity. In festival land, people often arrive for finite chunks of time and understandably repeat the systems they inherited. That can keep things moving in the short term, but it also means that small inconsistencies accumulate from year to year.
As Morgan Norwich humorously observed, SummerWorks is often “just two raccoons in a trench coat” before the seasonal team joins. The joke speaks to a reality shared by many arts organizations: when resources are stretched thin, building sustainable data systems can feel secondary to simply keeping the work moving forward.
SummerWorks approached this by documenting not only the dashboard, but the steps required to feed it. Their process included instructions for what to pull, from where, and how to format it. That means a future patron services manager, intern, seasonal staff member, or contractor does not have to reinvent the process from memory. The system becomes less dependent on one person’s brain.
TRANSFERABLE TIP: Every recurring data task should have a handoff note. At minimum, document the data source, export steps, fields to keep, fields to ignore, naming conventions, folder location, and the person responsible. This does not need to be elaborate. A simple “how to pull and clean this report” guide can protect organizational memory and reduce future stress. Don’t hate us for suggesting this, but AI can make this somewhat monotonous process significantly less labour intensive. Rather than starting from a blank page, staff can use AI to transform rough notes, screen recordings, or even a completed workflow into a clear, step-by-step procedure. As tasks evolve, those instructions can be quickly updated and refined.
Make data cleaning part of the practice, not an afterthought
The conversation also made clear that data cleaning is not clerical busywork. It is where much of the interpretive labour of data practice happens. Names, cities, pronouns, accessibility information, genres, and open-text survey answers can all be entered in inconsistent ways. The way an address is saved in one program is likely different than in another, making communication between two data sets incompatible without cleaning. Multiple-choice questions with “select all that apply” responses may need to be structured differently than single-choice questions. Reports exported from different systems rarely line up neatly on their own.
SummerWorks used AI-assisted prompts to help clean exported data, to save a person the time of manually reviewing and consolidating hundreds of thousands of data points. The important practice was not simply “using AI,” it included documenting the prompts, reviewing the results, and adding new cleaning instructions as unexpected patterns appeared in both the cleaned data sets and the data visualizations. Privacy and security were top of mind when choosing which AI tools to use for sensitive data.
TRANSFERABLE TIP: Treat data cleaning as a repeatable workflow. Keep a prompt bank or cleaning checklist. Record the transformations you make. Note decisions about categories, spellings, date formats, and multi-select fields. When possible, improve the original form or survey so that next year’s data arrives cleaner at the source.
Collect evidence during the rush, not only after it
PuSh focused on a challenge that many producing and presenting organizations will recognize: by the time the event is over, everyone is exhausted, and the details have already started to disappear. During the height of their festival, PuSh could have several events in a single day. The people who knew the most about a specific event were not always the same people responsible for year-end reporting or long-term analysis.
To address this, PuSh developed internal questionnaires and forms for staff and venue representatives to complete in real time. These captured attendance information, event experience, and qualitative observations while the information was still fresh. This allowed the team to gather day-by-day evidence rather than relying only on memory during a post-festival debrief.
TRANSFERABLE TIP: Build lightweight data collection into delivery. A short form completed at the end of each event can capture attendance notes, audience response, access issues, artist or partner feedback, operational challenges, and memorable moments. The form should be easy enough to complete during a busy day and structured enough to compare across events later.
Combine numbers with context
A recurring theme in the exchange was that numbers alone rarely tell the full story. SummerWorks was interested in how different communities intersect: ticket buyers who donate, artists who also attend, subscribers who submit, volunteers who become patrons, and so on. PuSh was gathering quantitative information about attendance alongside qualitative observations about experience.
This combination is especially important in the arts, where impact often lives in relationships, pathways, and accumulated trust. A single attendance number might tell you how many people came. It does not tell you whether an artist’s community showed up, whether a partnership deepened, whether an access measure worked, or whether an audience member is beginning a longer relationship with a particular artist that was showcased.
TRANSFERABLE TIP: Pair every key metric with a context question. If you track attendance, also ask what shaped attendance. If you track donations, ask what other forms of engagement donors have with the organization. If you track accessibility requests, ask what the organization learned about planning, resources, and care. This turns reporting into learning.
Give people access to insight, not necessarily every system
One useful distinction that emerged in the conversation was the difference between access to data and access to systems. Not everyone in an organization needs to work directly inside the CRM, ticketing platform, or raw spreadsheets. In fact, too many people editing source data can create confusion. But many people do need access to the insights those systems produce.
A dashboard or visual report can help marketing, development, programming, leadership, and board members understand relevant patterns without requiring everyone to manipulate the source system. This also supports better permissions and cleaner workflows. People can see what they need to see, while the underlying data remains protected and consistently managed.
TRANSFERABLE TIP: To help organizations tackle this challenge, Mass Culture will soon be sharing its own Data Governance Guide: a practical, adaptable resource developed through our work with arts organizations across the country. Rather than starting from scratch, organizations will be able to borrow, adapt, and build on the structures, templates, and approaches that Mass Culture has developed for organizing data responsibilities, workflows, and reporting practices. The goal is simple: make it easier for organizations to establish clear, sustainable systems for getting the right information to the right people at the right time.
Let messy history become a baseline
Both festivals were honest about the limits of their historical data. Some years were more complete than others. Platform changes made older data harder to compare. Pandemic years disrupted patterns. Certain kinds of patron or attendance information were only reliable from a particular point forward.
This is the normal starting point for many arts organizations. The important move is to name the limits clearly and begin creating a stronger baseline from today onward. Once an organization knows which years are reliable, which data points can be compared, and which gaps need to be acknowledged, it can still learn meaningfully over time.
TRANSFERABLE TIP: Create a “data confidence” note for each major data set. Label what is complete, partial, missing, or not comparable. This helps prevent overclaiming and gives future staff a clearer understanding of what the data can and cannot support. This practice was introduced to us by Maia Pelletier of Purpose Analytics during Mass Culture's 2025 Data Summer School.
Use AI as an assistant, not an authority
As the conversation unfolded, AI emerged not as a future possibility, but as a practical tool already being integrated into everyday data work. Participants shared examples of using generative AI to clean and standardize data, generate data-cleaning prompts, summarize qualitative feedback, identify patterns across datasets, and support the creation of visualizations. At the same time, they were grappling with important questions about privacy, security, and what types of organizational data should—or should not—be shared with AI tools.
Working with PuSh Festival, data coach Meena Das helped the team explore the strengths and limitations of different generative AI platforms and identify which tools were best suited to specific tasks. One example was Napkin AI, a tool designed to transform ideas, notes, and information into clear visual representations, making it easier to communicate insights across an organization.
AI offers significant potential to reduce repetitive work and accelerate analysis. But one of the clearest lessons from the exchange was that AI is most effective when it is built on a strong foundation of data practice. Clean data, documented workflows, thoughtful prompts, human oversight, and clear governance remain essential. AI can help reduce friction in the process, but it cannot replace the organizational systems and judgment that make data meaningful in the first place.
TRANSFERABLE TIP: A great place to begin exploring AI is the free nonprofit-focused tools offered by Logical Outcomes. They provide a free transcription service, a prompt library and and an evaluation tool. Although the organization plans to sunset its AI Workspace at the end of the summer, we still encourage arts organizations to take advantage of this Canadian-hosted, privacy-conscious platform while it is available. Logical Outcomes has earned a strong reputation for its transparency, ethical approach, and clear communication about how AI systems operate and how user data is managed—making it an accessible and trustworthy entry point for organizations beginning their AI journey. https://www.logicaloutcomes.net/ai-tools/
Share learning across organizations
One of the most exciting parts of the exchange was not a specific tool or technique. It was the recognition that organizations facing similar challenges can learn faster together. They were genuinely curious and wanted to know what one another was collecting, what systems were working, where the pain points were, and what trends might be visible only when organizations compared experiences.
This has implications beyond festivals. Arts organizations often work defensively around audience information, partner data, and internal processes. Yet shared learning can strengthen the whole sectoral landscape. If several organizations are asking similar questions about audience retention, access, donor engagement, artist pathways, or regional attendance patterns, there may be value in aligning some data practices and comparing lessons annually.
TRANSFERABLE TIP: Transferable tip: Consider convening a small group of peer organizations once a year to discuss what your data is teaching you. Start by sharing questions, definitions, indicators, dashboards, survey design, and lessons learned. Over time, this kind of exchange can strengthen data practices across a sector and create a stronger foundation for collective advocacy. One of the key lessons from the SummerWorks - PuSh exchange is that some of the most valuable learning happens not from the data itself, but from conversations about how organizations are collecting, interpreting, and acting on that information. Convening these kinds of conversations is also part of Mass Culture's role. As a national arts service organization, we see tremendous value in creating spaces where organizations can share approaches, learn from one another, and build collective capacity around data. If bringing together a community of practice around a shared data challenge would be helpful, we encourage you to reach out to us.
Advocate for data access in partnerships
The exchange also surfaced a sector-wide issue: organizations do not always have access to the box office or audience information connected to their own work, especially when they are presented by, produced within, or partnered with another organization. For smaller companies, this can be especially harmful. Without access to meaningful audience data, they may struggle to understand who they reached, follow up with communities, report accurately, or make a stronger case to funders.
Privacy, consent, and anti-spam rules matter. But they should not automatically prevent thoughtful data-sharing. Partners can clarify what information is collected, what consent language is used, what can be shared, and how data will be stored and used responsibly.
TRANSFERABLE TIP: Data access is not an administrative detail; it is part of equitable partnership design.When presenting organizations do not have access to audience, attendance, or ticketing data, their ability to understand, serve, and advocate for their communities is diminished. Equitable partnerships should include clear conversations about what data will be shared, how it will be shared, and how it can be used to support collective learning and sector growth. We encourage organizations to raise this issue through arts service organizations such as the Professional Association of Canadian Theatres (PACT) and other sector networks to help move the conversation forward.
A starting checklist for arts organizations
For organizations wondering where to begin, the exchange suggests a practical starting point. You do not need a perfect CRM, a custom-built dashboard, or a full-time data analyst to improve your data practice. You need a manageable system that helps people ask better questions and carry knowledge forward.
Start with these questions:
- What are the three most important stories we need our data to help us tell?
- Where does the relevant data already live?
- Who owns each data source, and who needs the insights from it?
- What needs to be cleaned or standardized before the data is useful?
- Which tasks recur every year, season, or project, and are they documented?
- What can be collected during delivery, before memories fade?
- Which historical data is reliable enough to compare, and where do we need a new baseline?
- What data-sharing expectations need to be clarified with partners?
- What parts of the process could be supported by AI, and what privacy safeguards are needed?
- Who else in our sector is asking similar questions, and what could we learn together?
Data practice as organizational care
The SummerWorks and PuSh exchange reminded us at Mass Culture that data practice is not separate from artistic practice, access practice, producing practice, or community practice. It is one of the ways organizations remember, learn, and make decisions with greater care.
Healthy data systems need to be usable. They need to reflect the realities of capacity, turnover, seasonal urgency, and imperfect information. They need to help people see patterns without losing the stories behind them. And they need to make it easier for the next person to pick up the work.
For festivals, that means carrying knowledge from one intense cycle into the next. For the broader arts sector, the lesson is just as relevant: the most useful data is not the data we collect once. It is the data we are able to understand, share, question, and return to over time.