ENTERPRISE B2B | MULTI-TENANT UI | TRANSACTIONAL DESIGN
ENTERPRISE B2B | MULTI-TENANT UI | TRANSACTIONAL DESIGN
Enterprise hospitality brands were losing sales volume due to a multi-vendor software ecosystem that stalled pipelines and bloated operational overhead. As the Lead Product Designer, I designed a native booking application and back-of-house portal that streamlined reservation workflows into a single scalable solution.
If you have limited time, just read the → arrows for a 60 second summary of my strategy & product impact.
PROJECT OVERVIEW
PROJECT OVERVIEW
Fragmented, multi-vendor booking systems stalled guest sales pipelines and cost enterprise hospitality providers significant revenue. The business needed a native, embedded reservation platform to secure market share and seamlessly replace entrenched legacy competitors without disrupting active operations.
I led end-to-end product design and user experience strategy of an embedded booking platform that allowed hospitality providers to consolidate their operations into a single ecosystem, successfully unifying and stabilize high-volume enterprise pipelines.
The primary strategic goal was defensive product consolidation. The initial scope required anchoring the end-to-end UX architecture for two environments: a highly flexible consumer-facing booking wizard embedded natively onto client sites, and the complex back-of-house management portal used by operators to configure varied inventory types.
Successfully displaced an entrenched niche competitor to secure single-vendor account exclusivity.
Shielded the product flow inside a Shadow DOM web framework, ensuring custom client implementations never suffered.
Brought highly diverse resort assets ranging from concert ticket sales to golf cart rentals into a single, cohesive user experience.
STRATEGIC CHALLENGE
Enterprise resorts were caught between two platforms. Spark excelled at managing complex schedules, backend events, and digital endpoints. Conversely, a niche competitor owned the transactional booking and reservation flows.
Operating both platforms introduced severe operational friction for resort staff and fragmented the guest experience. Clients demanded a single vendor. To protect its ecosystem and expand market share, Spark needed a native reservation engine capable of handling any kind of bookable experience: from hourly cabana rentals to fixed-schedule yoga classes and high-volume concert ticketing.
SYSTEM ARCHITECTURE
Designing a platform that converts raw resort inventory into a seamless purchase requires balancing consumer accessibility with complex backend operational realities. I architected the end-to-end interface logic across 6 distinct user profiles:
The Planner (Pre-Arrival): Guests browsing available experiences months in advance, requiring information layouts and grid views to plan their itineraries.
The Browser (On-Site): Guests wandering the property, interacting with resort communications, requiring scannable, high-contrast layouts optimized for quick decision-making.
The Active Booker: Guests committed to purchasing, moving through a high-intent transactional checkout wizard.
The Concierge: Staff member assisting a guest live or over the phone, needing rapid bypass options to book experiences, apply manual overrides, or alter reservations on the fly.
The Third-Party Host: External vendors (e.g., local charter captains or yoga instructors) who need isolated views to monitor their daily rosters without accessing sensitive resort analytics.
The Operator/Manager: Power users responsible for creating, defining, pricing, and managing the entire matrix of bookable assets inside the backend system.
UI STRESS-TESTING AGAINST THE SHADOW DOM
The business model dictated that the booking flow work perfectly as a standalone web page and as a white-labeled experience embedded via shadow DOM directly onto client websites. This meant our components had to be completely isolated from host site CSS collisions while remaining flexible enough to inherit their visual branding parameters seamlessly.
To ensure visual integrity across variable deployments, I established a step-based modular layout strategy. We isolated and quarantined recurring structural elements, and broke the dynamic data collection into a step-by-step wizard.
To defend against downstream layout breaks, I used automated AI design pipelines during the wireframe exploration and high-fidelity crafting phases. Leveraging standardized inputs via Google Stitch and DESIGN.md files, I systematically swapped out various brands to spot check layout success.
This allowed us to stress-test the UI across simulated resort brand profiles before development sprints began, eliminating structural layout bugs caused by unpredictable host-site styling.
TACTICAL EXECUTION & DELIVERABLES
The front-end layout had to elegantly handle different booking and inventory constraints within a unified flow. For example, a concert booking requires priority for date and time selection, a golf cart rental depends on duration scaling, and a cabana booking requires zone mapping. I designed a flexible UI approach that adapts to the inventory type chosen by the operator.
Architected by me, but developed by the lead engineer: Resort bookings suffer from high friction. Guests often abandon carts when researching and coordinating their experiences. To combat this, I designed a shopping cart that seamlessly allows users to view reservations that they closed halfway through the booking process. The interface leverages local storage to securely snapshot and cache guest input progress locally. If a guest closes their tab or loses connectivity, the system prominently displays in-progress bookings, allowing them to instantly resume their reservation right where they left off.
MEASURABLE OUTCOMES & RETROSPECTIVE
Resolved key feature demands for 50% of active enterprise accounts, transforming a primary sales bottleneck into a high-value upsell opportunity.
Enabled flagship properties like Islamorada and Grand Hotel to cleanly terminate their subscriptions with competing platforms, consolidating their entire digital operation under Spark.
Proactive edge-case stress testing and automated style variations neutralized downstream layout technical debt, resulting in zero layout regressions reported post-deployment.
This platform proved that B2B SaaS design cannot exist in a vacuum. To build a successful transactional solution, I had to balance the isolated constraints of front-end web development with the complex realities of resort operations. Success meant more than just designing pretty interfaces; it meant creating an ecosystem robust enough to outcompete an established market player.