About

Envisioned an AI-enabled, natural language search experience that curates content from across Marriott’s hospitality portfolio and serves results in a predictive, personalized and differentiated way for users.

Team

Gen AI Deloitte Leadership
Deloitte Principals & Managing Directors
Client Stakeholders (C-Suite)
Product Management
Creative Director
Motion Designers

Deliverables

Creative Visioning
UX/UI Design
Prototypes
Motion Design

Understanding the brief

When Marriott came to Deloitte to get our POV on the future of AI powering their booking search results, I was engaged to help concept ideas, guided by questions like:                                                                                                        

How might we...

“…use AI to enable freeform text queries?”
“…better personalize the search results?”
“…get users to search more visually?”

Grounding a freeform, aspirational ask

To get a handle on the direction, I asked clarifying questions to guide my thinking:

Q: Design requirements?

A: Must feel like Marriott and stay true to the brand

Q: Technical constraints?

A: Understand the basics around Large Language Models (LLMs) & Natural Language Search (NLS)

Q: Are there any guardrails around functionality?

A: Think big. Keep AI and future LLMs at the center of powering the experience

Q: What’s the scope?

A: Focusing solely on the search and search results page experience

Gathering inspiration

To jump start ideas, I look across the industry and beyond to get an idea of how other brands are approaching similar concepts. Next, I tinker with possibilities on how to improve what is in-market today.

I documented my findings in a FigJam for easy collaboration across the team and to build artifacts we can reference throughout our process.  

Outlining initial concepts

In order to generate ideas and feedback quickly, I opted to sketch ideas on FigJam. This allowed me to show progress quickly and put the concepts in front of my Creative Director and Deloitte leadership and get validation that I’m heading in the right direction.

In our initial approach, we thought about how different traveler personas are searching for their hotel stays and how might “faceted search” help enhance that experience:  

Taking in feedback

After demoing these concepts with our internal stakeholders, we were given the following feedback:

→ Center the experience around gaining inspiration

→ Make the experience more personal for users

→ Break out of the traditional hotel search results box

→ Inspire Marriott-driven experiences, beyond rooms

Pulling inspiration from the brand

When flipping through the Marriott brand guidelines I was struck by their “connector” element that is used throughout their branding and featured in the Bonvoy logo.

Leaning in on existing photography

Throughout the guidelines, I kept noticing the varied and aesthetic Marriott’s property photography.

The photography guidelines stated:

“Marriott’s tertiary photography aims to be from the POV of the traveler”.

This was to serve as the core, immersive element that I used to guide the vision of the next iteration.

Sketching a concept

I took pencil to paper this time to express how the connector could be leveraged as a UI element to toggle search results between “Inspiration” to “Rooms”, “Restuants” and more.

Wireframing

I then expressed my idea in more detail in a FigJam sketch to test out how search results could be varied in style and content.

I used this sketch to validate with the team, who agreed we were on the right track.  

Product Walkthrough

Delivering the vision through storytelling & prototyping

Our Users

Our sample users, Micah and Cailey, who prioritize an adventurous travel and are open to using Natural Language Search.

MICHAH’S STORY

Micah is authenticated Bonvoy member. The system has learned about his interests and shows his interest in architecture.

Micah and Cailey’s first date was seeing the movie Eat, Pray, Love and the two have discussed Italy for their honeymoon. He uses these ideas to begin his search.

He’s consider’s himself a foodie, and adds that prompt to his search.

PRODUCT NOTES

I wanted to encourage the user to add rich details to populate custom results and developed these dynamic interest chips that can be added to the query in a single-click.

This vision allowed us to reimagine the homepage hero and the search bar to allow for and promote Natural Language Search, powered by AI

MICHAH’S STORY

Micah explores the moodboard for inspiration on planning Cailey and his dream honeymoon.

PRODUCT NOTES

Powered by AI, the system recognizes that this is an inspiration-driven search and auto toggles to the Insipiration mood board.

MICHAH’S STORY

Users can scroll and explore Marriott driven sites, stays, and trip ideas, pulling from Marriott’s vast content ecosystem.
They can easily save ideas for later and share with collaborators. (Move this statement)

Explore other ways of searching - by property, by dining options

PRODUCT NOTES

Powered by AI, the system recognizes that this is an inspiration-driven search and auto toggles to the Insipiration mood board.

MICHAH’S STORY

Micah saves his moodboard to his shared collection so Cailey can view it.

PRODUCT NOTES

Users can scroll and explore Marriott driven sites, stays, and trip ideas, pulling from Marriott’s vast content ecosystem.

They can easily save ideas for later and share with collaborators.

CAILEY’S STORY

Cailey gets a push notification that Micah has added more to their collection.

Seeing the new addition, she taps to view.

CAILEY’S STORY

She’s interested in a spa experiences on their trip and customizes the search.

CAILEY’S STORY

She explores her customized moodboard and saves her ideal spa locations.

CAILEY’S STORY

When they're ready, the couple selects their dates and books their itinerary.

The Final Vision

View the entire prototype recording

Presenting to Marriott

I represented our collective vision at Marriott HQ in front of their Product & Marketing Executives. We received positive feedback, which opened the door to new conversations around search, AI, and future partnerships with the brand.

Team feedback post-demo