Melissa ShiConsumer travel · Product design · Mixed-method research
Nemacolin Woodlands Resort Trip Planner
Turning more than 80 resort activities into a guided, personalized path from discovery to booking.
Problem
Nemacolin offered exceptional service and more than 80 activities, but prospective guests struggled to understand its unique value and did not reserve enough of the experience before arrival.
Solution
A responsive trip planner uses guest goals and trip context to recommend stays, activities, and dining while preserving control through an editable itinerary.
Impact snapshot
Five rounds of evidence moved the product from novelty to conversion
The final direction was not a single idea taken to high fidelity. Each research round resolved a different product question: who to design for, where personalization added value, how much structure guests needed, and what would actually encourage action.
Research foundation
The visible conversion problem started before booking
Interviews with past guests, a mystery booking exercise, market analysis, reservation-data analysis, and onsite contextual inquiry showed that guests arrived without enough planning. Staff could not reliably promote the resort's full activity portfolio, and missing reservations could diminish the experience before a guest reached the property.
Activities were the differentiator
Nemacolin's breadth of experiences was its strongest advantage, but that value was difficult to communicate.
Pre-arrival action mattered
Reservations were essential to accessing the best experience, yet guests often delayed planning until arrival.
Mobile was the planning surface
Research participants commonly used phones for trip discovery and planning, supporting a mobile-first direction.
Decision 1
Target motivations instead of demographics
Reservation data initially pointed to a target segment of upper-middle-income adults ages 35 to 45, often traveling with children. Contextual inquiry revealed a more useful segmentation variable: why people were taking the trip.
What if recommendation logic began with the value a guest wanted from the trip, not only who the guest appeared to be?
Connect motivation to perceived value
Across couples, families, and friend groups, motivation changed what guests noticed, trusted, and prioritized. I reframed the experience around active, adventurous, cultural, relaxing, romantic, and family-oriented goals.
Use behavioral data as supporting evidence
Historical reservation, CRM, call-center, and feedback data identified eight useful booking signals. The model reached 43% hotel-classification accuracy, useful for narrowing options but not strong enough to replace stated preference.
The resulting strategy combined explicit trip goals with contextual signals instead of treating demographic similarity as personalization.
Decision 2
Support the full reservation journey
The first concepts explored an interactive room, a recommender, and gift giving. I designed speed-dating research with prospective guests and resort staff to validate the underlying needs before the team committed to a solution.
Test needs before polishing features
Storyboards made each value proposition easy to compare without investing in detailed interface work.
Expand from recommendation to completion
Concept validation showed that a useful solution needed to connect resort discovery, option comparison, itinerary planning, and payment. A recommendation-only feature would have stopped before the hardest decisions.
Decision 3
Choose flexible planning over a prebuilt answer
We tested two ways to encourage reservations: let guests build an itinerary while booking, or present a ready-made itinerary. I planned guerrilla, in-person, and remote sessions to compare the mental models despite recruiting constraints.
Compare complete workflows
The first flow offered more control; the second reduced effort. Testing showed that guests valued flexibility, richer option details, and clarity about which days recommendations applied to.
Carry the strongest parts forward
I recommended the build-your-own direction, then added detailed stay and experience information, easier itinerary management, and a day-based view.
Decision 4
Add structure without removing control
Iteration two exposed a different problem: flexibility alone left guests unsure what the planner was for and how to move from interest to booking. I reframed navigation around a visible four-step journey.
Quiz
Capture party, dates, and trip style with a focused sequence.
Recommendations
Explain relevant options while keeping alternatives visible.
Itinerary
Organize choices by day and preserve editing flexibility.
Booking
Carry selected stays, activities, and dining into one completion path.
Later tests confirmed that participants understood the workflow and could move through it smoothly. Those sessions also revealed secondary cases such as returning to a saved itinerary, handling availability conflicts, and booking a room without activities.
Decision 5
Use personalization without overstating the model
The final recommendation logic filtered a large catalog using party composition, travel timing, stated trip style, and historical patterns. Because first-time guests had no prior behavior, the underlying model classified likely hotel preference rather than operating as a true recommender system.
Keep the algorithm legible
The quiz collected only signals that changed results, while recommendations retained visible context such as dates, party, trip style, and price.
Use persuasion with guardrails
Ten design principles balanced conversion goals with guest needs: transparency, flexibility, clear expectations, scannable choices, relevant scarcity, and a modern expression of white-glove service.
Final experience and results
A guided planner that made a complex resort feel easier to choose
Conversion intent
All 25 participants said they were likely to book a stay if price were not a constraint.
Earlier reservations
Ninety-two percent were likely to reserve ahead, and 42% said they would book earlier than usual.
Brand perception
Participants described the experience as modern, professional, efficient, organized, helpful, clean, inviting, and easy.
My biggest takeaway was that personalization is a product strategy, not an algorithm alone. The strongest outcome came from combining behavioral evidence with stated intent, then giving guests enough structure to act without taking away their control.
Keep exploring