Case Studies
See how our approach has been applied to analyze stakeholder input, develop programs, and understand issues. Learn about the organizations and funders supporting this work.
Case Studies
Analysis of barriers to arts participation from a national survey on arts engagement, identifying key themes around access, awareness, and personal constraints.
Key Findings:
- •Time constraints emerged as the dominant barrier across demographics
- •Cost concerns varied significantly by geographic region
- •Awareness gaps highlighted opportunities for community outreach
Community input on local issues and priorities for the City of Ann Arbor, helping planners understand resident concerns and suggestions.
Key Findings:
- •Transportation and mobility ranked highest in resident concerns
- •Housing affordability showed strong generational differences
- •Environmental sustainability connected to multiple topic areas
Topic modeling of Detroit 311-style citizen reports to surface neighborhood concerns: infrastructure, waste management, safety, and quality-of-life issues from resident-submitted requests.
Key Findings:
- •Yard management, drainage, and waste disposal emerged as a major cluster of property-level concerns
- •Urban infrastructure—streetlights, sinkholes, fire hydrants, tires—formed a distinct theme with strong safety implications
- •Resident narratives revealed inequities in service response and enforcement alongside recurring location-specific issues
Public comments from Michigan municipal meetings (Ge et al., 2025)—multi-sentence civic remarks on housing, services, safety, and local governance, with city-level metadata for comparison across places.
Key Findings:
- •Broad thematic structure spans planning, accountability, and neighborhood-level concerns voiced in public forums
- •Multi-city coverage supports comparing how themes show up across municipalities when city metadata is present
- •Hierarchical topics and an outliers view help surface both dominant narratives and long-tail perspectives
Analysis of invention abstracts to understand research themes and identify emerging areas of innovation across university portfolios.
Key Findings:
- •AI/ML applications appeared across traditional disciplinary boundaries
- •Healthcare innovations showed strong clustering patterns
- •Sustainability research connected multiple domain areas
Our Approach
Our methodology builds on these key activities:
Open-ended responses are gathered through surveys, interviews, public consultations, or other engagement methods to capture stakeholder perspectives and concerns. Text responses are then cleaned and prepared, removing duplicates and empty entries while preserving the original voice.
BERTopic-based modeling discovers natural groupings in the data without predefined categories. Topics are then organized into tree structures, with LLM-generated labels and summaries at each level.
Build shared understanding around analytical results, using visual hierarchies and summaries as common reference points that help groups move beyond anecdotes toward grounded discussions.
Situate analytical results within specific organizational, policy, or design contexts, mapping themes onto existing goals and identifying where strategies align or diverge from expressed concerns.
Use outputs to anchor planning sessions, design workshops, and strategy meetings, enabling activities like prioritizing issues, identifying leverage points, and stress-testing proposed actions against the distribution of concerns.
Support iterative use by tracking shifts in attention and concern over time, assessing intervention effectiveness, and enabling systems of learning rather than one-off studies.
Partners
Funders
Supporting arts engagement research and public participation initiatives.
Funding research on humanities applications and public scholarship.
Supporting civic engagement and democratic participation research.
Collaborators
Academic home and primary research institution for the project.
Collaborative research and development partner for the project.
Intellectual property and commercialization partner for the project.
Work with us
The organizations behind these case studies share a common problem: more input than they could meaningfully act on. If that sounds familiar, we'd like to talk.
Voxil is well-suited to:
- Public agencies and municipal governments. Running community input processes, participatory budgeting, or planning consultations where resident voice needs to inform real decisions — and where the scale of input defeats conventional analysis — and where transparency builds trust.
- Foundations and nonprofits. Conducting needs assessments, program evaluations, or stakeholder consultations where the diversity of perspectives matters as much as the headline findings — and where funders and boards need to see that the full range of voices was heard.
- Research institutions. Studying public opinion, civic engagement, or community concerns at scales that conventional qualitative methods can't reach — where interpretive accountability to source material is a methodological requirement, not a nice-to-have.
We work closely with partners through the full process — from data preparation through analysis, interpretation, and activation. Every engagement is different.