Visitor research reduces the risk of building an elegant answer to the wrong problem. Before final scripts and recordings, study how people actually arrive, navigate, look, listen, decide, ask for help, and experience barriers.
Start small but real: review existing evidence, observe visits, speak with visitors and staff, map key journeys, and test a rough playable prototype. A few grounded research rounds are more valuable than a detailed persona invented in a meeting.
1. Turn assumptions into research questions
Write decisions the project must make: Which route is understandable? When do visitors want optional depth? What prevents audio use? Which languages and access formats are unmet? Can families listen together? Does the first story direct attention effectively?
Avoid questions designed to validate the proposed solution, such as “Would you like an innovative AI guide?” Ask about behavior, context, needs, and trade-offs.
2. Review existing evidence
Combine attendance and ticket data, previous surveys, complaints, staff observations, accessibility audits, website searches, program participation, language requests, and existing digital analytics. Record where evidence is old, incomplete, or unable to represent non-users.
3. Observe the real visit
Observation reveals what people may not remember or mention: missed entrances, crowding, sound bleed, unread labels, scanning posture, group negotiation, screen distraction, and early fatigue. Use an ethical observation protocol and avoid recording identifiable information without an appropriate basis.
Note context, action, apparent barrier, and a possible follow-up question. Do not label motivation from behavior alone.
4. Interview without leading
Ask visitors to describe what they expected, chose, found difficult, enjoyed, or would change. Use neutral prompts: “Tell me what happened when you entered,” “How did your group decide where to go?” and “What made you start or not start the guide?”
Include non-users, short visits, repeat visitors, different languages, families, school groups, and people with relevant access needs. Front-of-house staff provide a separate operational perspective.
5. Map journeys and audience contexts
Create evidence-based journey maps for important situations rather than one average visitor. Include pre-arrival information, companions, time, language, confidence, device, access, entrance decisions, route transitions, rest, exit, and continuation.
Personas are useful only when they summarize evidence and remain revisable. Avoid stereotypes based on age, nationality, or disability.
6. Prototype before production
Test the cheapest representation that can answer the question: printed route cards, clickable screens, temporary QR codes, a staff-read script, or a rough voice recording. Formative testing should happen while change is still affordable.
Ask participants to use the prototype rather than only discuss it. Observe start, navigation, looking, listening, comprehension, group interaction, and recovery.
7. Include access needs from the first round
Recruit people whose sensory, motor, cognitive, language, and digital contexts are relevant. Compensate specialist and community contributions appropriately. Accessibility research is not a substitute for standards review, and standards review is not a substitute for user research.
8. Translate findings into decisions
For each finding, record evidence, confidence, affected audience, decision, owner, and what should be tested next. Distinguish a repeated pattern from an individual preference without dismissing an issue that may cause serious exclusion.
9. Where GuideSofia fits
GuideSofia can support a pilot using real content, route choices, languages, and access options before full production. After launch, privacy-conscious engagement signals can complement continued interviews and observation.
10. Visitor-research checklist
- Research questions connect to real decisions.
- Existing evidence and its limitations are documented.
- Observation covers real visitor and venue conditions.
- Users and non-users are included.
- Relevant access and language contexts participate.
- Prototypes are tested before expensive production.
- Findings distinguish evidence, inference, and recommendation.
- Decisions, owners, and follow-up tests are recorded.
Frequently asked questions
How much visitor research is needed before an audio guide?
The scale should match the decision and risk. Even a small project can observe real visits, interview a diverse set of visitors and staff, and test a rough prototype before producing the full guide.
Can analytics replace visitor interviews?
No. Analytics show interface behavior but usually not motivation, physical context, comprehension, exclusion, or emotional response. Combine behavioral and qualitative evidence.
Should a museum use visitor personas?
Personas can help if grounded in evidence and treated as decision tools rather than stereotypes. Include context such as time, companions, language, confidence, access needs, and visit motivation.
When should accessibility testing happen?
From the first prototype onward. Late testing may identify structural barriers after routes, interfaces, media, and procurement decisions have become expensive to change.