Applied UX Research · Q1 2026 · Tourism AI

Testing an AI Trip Planner for International Tourism Audiences

Moderated usability testing and user interviews evaluating a tourism website and its embedded AI trip-planning tool with international prospective travelers.

Role Product Research Consultant
Method Moderated usability testing
Scale 12 sessions led; 40+ wider interviews
Impact Mobile, category, and AI trust priorities surfaced
01 — Snapshot

Project Snapshot

Role
Product Research Consultant
Timeline
Q1 2026
Product Type
AI Trip Planner
Embedded in tourism website
Participants
International prospective travelers
Confidentiality
NDA — Client Anonymized
12
Moderated Sessions Led Personally
40+
Interviews Across the Wider Study
6
Journey Stages Mapped
8
Core Friction Themes Identified
What I did
Led moderated sessions and observed screen-share behavior across the trip-planning journey.
What surfaced
Mobile blockers, category confusion, information overload, and provisional AI trust.
What changed
Findings informed bug escalation, category restructuring, and AI UX roadmap priorities.
02 — Context

Overview

Evaluated a tourism website and its embedded AI trip planner with international prospective travelers, observing the full journey from discovery to travel decision.

My Role
  • Product Research Consultant
  • Usability Researcher
  • Session Moderator
Research Activities
  • Moderated Usability Testing
  • Screen-Share Observation
  • Think-Aloud Protocol
  • Post-Task Interviews
Outcomes at a Glance
  • Mobile blockers escalated & tracked
  • Category restructuring recommended
  • AI trust factors surfaced for product team

Context

Platform Campaign Website + Embedded AI Trip Planner
Audience International Prospective Travelers
Focus Discovery → Decision Journey
Primary Goal Inspire and convert travel intent

Challenge

  • Category comprehension across cultures
  • AI planner trust & usefulness
  • Mobile & technical barriers
  • Cultural fit of content organization

The website was designed to inspire international travelers — but whether its structure, content, and AI planner actually worked for non-domestic audiences had never been tested.

Study scope: This case study covers the qualitative and observational track of a larger mixed-method study. Behavioral observation revealed friction points that the survey component alone could not surface.
03 — Research

Research Focus

Objectives and guiding questions across the moderated sessions.

01
Category Comprehension
Assess whether international users understand and can navigate the site's category structure.
02
AI Planner Expectations
Understand when, why, and with what expectations users open the AI trip planner.
03
Trust & Personalization
Evaluate whether AI outputs feel personal and trustworthy enough to influence travel decisions.
04
Mobile Barriers
Identify bugs and interaction barriers that block task completion on mobile.
05
Content Confidence
Locate where information overload or missing content reduces decision confidence.
03 — Research

Session Flow & Methods

Each moderated session walked participants through the full journey — from discovering the site to deciding whether to act on an AI-generated plan.

1
Discovery
How did they find the site?
2
Category Navigation
Category comprehension & clicks.
3
Page Review
Content clarity & information load.
4
AI Planner Entry
When & why they opened it.
5
AI Output Review
Trust, usefulness, personalization.
6
Decision Confidence
Would they act on it?

Methods

Behavioral & Qualitative
  • Moderated Usability Testing
  • Screen-Share Behavioral Observation
  • Think-Aloud Protocol
  • Post-Task Interviews
Context & Content
  • Category Comprehension Testing
  • AI Planner Interaction Testing
  • Mobile Usability Review
  • Survey Context (wider study)
04 — Analysis

Analysis & Findings

Friction mapped across the full journey from discovery to decision.

Stage Friction Observed Severity
Discovery Many participants did not know the website existed — organic discovery and brand recognition were low for international audiences. High
Category Navigation Destination-specific category names rooted in domestic conventions confused international users; overlapping categories created uncertainty. High
Page Review Long scrolling and dense information caused anxiety and overload — participants skimmed or gave up rather than reading thoroughly. Medium–High
Content Organization Content was rich — depth was even praised once found in collapsible sections — but it felt scattered across long pages. The disorganization pushed users toward the AI planner, which they preferred for its perceived personalization. Medium
AI Planner Entry Users needed clearer cues for when and why to open the AI planner; many missed or delayed its entry point. Medium
AI Output Review Users liked the AI's answers — but trust was provisional, and one wrong detail could discredit everything. High
Mobile Use Stuck-page bugs blocked task completion on mobile — the most critical technical finding of the study. Critical
Key Insight

Users liked the AI's answers — but trust was provisional, and one wrong detail could discredit everything.

05 — Outcomes

Outcomes

Research contributions delivered to the client's product and content teams.

Deliverables

Field Notes
Session Observation Records

Behavioral observation notes for every moderated session.

Issue Log
Usability & Bug Reports

Friction points catalogued by severity and journey stage, including mobile reproduction steps.

Strategy
Category & Content Recommendations

Restructuring guidance to make navigation intuitive for international audiences.

Report
Research Synthesis

Findings and AI planner UX recommendations delivered to the client team.

Impact

Mobile Blockers Escalated
Stuck-page and interaction bugs logged, escalated, and tracked to resolution
Category Restructure Informed
Overlapping and confusing categories identified for restructuring by the content team
AI UX Priorities Set
Entry-point and trust factors surfaced for the product team's roadmap
06 — Reflection

Reflection

What this project revealed about researching AI features and cross-cultural audiences.

What Worked
Behavior Over Self-Report

Screen-share observation revealed mobile blocks, delayed AI planner entry, and overload responses that post-task surveys alone would have underrepresented or missed entirely.

What Surprised Me
Mental Models Don't Travel

Category and navigation logic that felt natural to domestic designers broke down for international users — assumptions didn't translate across cultural contexts.

Key Takeaway
Cultural Fit Must Be Tested, Not Assumed

For destination and tourism content aimed at international audiences, comprehension and trust have to be validated with the actual audience — not inferred from domestic intuition.

Back to Research Portfolio

View more applied UX, product, and marketing research projects.

← View All Projects