All case studies
UX Design Booking Engine Theming

Redesigning a booking engine that has to fit hundreds of properties

Rebuilding Agilysys's online booking experience mobile-first, around multi-guest booking, smart availability, and a themeable base, so every resort, hotel, and spa can make it their own.

Agilysys (Book4Time)
Lead & sole designer
Research, UX/UI, Theming

A booking engine that couldn't scale across hundreds of properties, lost guests mid-flow, and was weakest on the phone, where most guests book.

As the only designer, I led research and redesigned the booking engine mobile-first: a themeable base, native multi-guest booking, and whole-visit availability.

Shipped as one adaptable foundation for hundreds of properties. Post-launch metrics are still to come.

One booking engine, hundreds of different businesses

Agilysys's spa and wellness platform has an online booking engine, where guests browse a property's offerings, pick a time, and book. It is embedded in the websites of hundreds of properties, from resort spas to hotel wellness centres to activity-led retreats.

No two of those properties look the same. Each has its own offerings, guest types, pricing rules, and brand. That is the real design brief: the booking engine can't be designed for one ideal property. It has to be adaptable enough to work for all of them.

It also has to work on the device guests actually use. Roughly 70% of guests book on a phone, so mobile is the primary experience.

A foundation that couldn't keep up with the product

As the platform gained functionality, the original booking engine became harder to extend, and the problems showed up for guests and for the clients who relied on it.

01

Hard to scale

Every new capability (packages, add-ons, activities, member pricing) had to be squeezed into a structure that was never designed for it.

02

Drop-off in the flow

Guests were abandoning bookings before completing them, and the friction in the flow was a major reason.

03

Dated, inconsistent UI

Screens had accumulated one-off patterns over time, which made the experience feel dated and made every change slower and riskier.

04

Weakest where most guests are

Roughly 70% of users book on mobile, yet mobile was the weakest part of the product: a desktop layout squeezed onto a small screen. Fixing it was the priority, and it called for a mobile-first approach.

Listening to the people who run the properties

Because the booking engine has to serve so many different businesses, I started with the businesses. I spoke with a wide range of clients about what they wanted, where they struggled, and what look and feel they were aiming for. I paired that with a competitor review and prototype testing.

Client conversations

Pain points, goals, and brand vibe

What clients struggle with in the current flow, what their guests ask for, and the look and feel each property is trying to achieve.

Stakeholder interviews

Operations and internal teams

How offerings, pricing, and guest rules are actually configured day to day, and where the product fell short.

Competitor audit

How other booking engines cope

A review of how comparable booking experiences handle multiple guests, availability, and customisation.

Prototype testing

Validating the flow early

Testing prototypes of the redesigned flow to pressure-test the booking logic before it was locked in.

Four requirements came out of this work and drove every decision after: the design had to be adaptable per property, booking for more than one guest had to feel natural, availability had to account for the whole visit rather than a single service at a time, and, because most guests book on a phone, the experience had to be mobile-first.

A flow built for one guest on one big screen

Before redesigning, I walked the existing flow end to end. It had been built around a single guest on a desktop, and every extra guest, service, or smaller screen exposed that. Client branding is removed below; the screens show a generic demo property.

Existing offering details page
Existing offering details
Existing availability page
Existing availability
Existing checkout with a repeated guest information form for each guest
Existing checkout: guest details repeat for each guest and service, producing a long form

Four decisions that shaped the redesign

All four come from the same idea: keep the structure consistent and let each property change the surface. Each one is tied to the problem it answers.

Final designs adjusted per NDA. Contact fionaqylai@gmail.com for the full case study.

Answers: Hard to scale · Dated UI

1. A base theme every property can make their own

Rather than design one look, I designed a strong, neutral base theme that works out of the box, plus a system around it that lets each client choose their own typography, colours, and imagery. The hard part was deciding where to draw the line: too much freedom and every property becomes a one-off to support; too little and it stops feeling like theirs.

Fixed for every property

Structure, spacing, interaction patterns, and booking logic. This keeps the experience reliable for guests and maintainable for the team.

Chosen by each property

Typography, colours, imagery, and how guests enter the flow, from the full menu to a "Book now" entry point. This is how each property expresses its brand.

The same thinking applied to content. Offerings come in different shapes: single services, regular packages, choice packages ("choose 3 offerings from options provided"), couples services, add-ons, and member pricing. I designed a small set of flexible offering templates that adapt to what a property sells, instead of a layout per case.

Offerings list with categories, services, packages, and Shop links
Fig 01 — Offerings list: services, regular and choice packages, and a couples service share one card system
Offerings list variant with a Book now entry point
Fig 02 — Variant with a "Book now" entry point for properties that want a different way in
Choice package offering details with per-choice selection
Fig 03 — Choice package template: each choice is picked from the options the property configures, with a running price
Answers: Drop-off in the flow

2. Multi-guest booking that doesn't multiply the work

Groups, couples, and families book together, and the old flow treated each additional guest as a separate chore. The redesign puts the guest at the centre: a "Booking for" switcher on every offering, add-ons chosen per person, and a choice between adding for the selected guest or for all guests. Member pricing is shown up front and applied at checkout.

Service offering with Booking for Sarah and Jordan, per-guest add-ons, and member pricing
Fig 04 — Offering details with a guest switcher, per-guest add-ons (Sarah adds Head spa, Jordan adds Facial therapy), and "Add for selected guest" / "Add for all guests"

Guest information follows the same logic. Each guest gets their own tab, a "Same as reservation primary contact" shortcut, and a sign-in prompt for returning guests, so the booker only types what's new. A running "Your visit" summary keeps every guest's services and price in view.

Guest information with a tab per guest and a Your visit summary
Fig 05 — Guest information: one tab per guest, shortcuts to avoid retyping, and a live visit summary
Answers: Drop-off in the flow

3. Smart availability that plans the whole visit

Guests think about their whole visit, not one service at a time. Availability now reflects the full itinerary: one view that shows every guest's services and add-ons, technician preferences up front, and a "same room" option for couples and groups. Times are priced as a whole, so guests aren't discovering conflicts after they've chosen. The result is fewer dead ends between picking an offering and confirming a time.

Choose a time page showing available times for multiple guests and services
Fig 06 — Availability for the whole visit: preferences on the left, times and prices for every guest and service on the right
Answers: Weakest where most guests are

4. Mobile-first, because most guests are on a phone

With most guests booking on mobile, I stopped treating the phone as a smaller desktop and designed for it first. Every flow starts as a single column built for thumbs: sticky actions anchored to the bottom of the screen, a bottom sheet for booking options and the running summary, and a collapsed "Your preferences" panel that expands only when guests need it. The desktop layouts then add space and side-by-side panels, rather than changing the structure, so guests get the same flow on any screen, and the mobile experience is no longer the weakest link.

Mobile availability for multiple guests
Fig 07 — Mobile availability: one column, collapsible preferences, and a sticky "Confirm"
Mobile offering details with booking summary
Fig 08 — Mobile offering details with the booking summary as a bottom sheet

Carrying every guest through to confirmation

Checkout lists the guests, the time, and an itemised visit per guest, with promo codes and gift cards applied in context and edit links back to earlier steps. The confirmation page then breaks the itinerary down by guest, with a reservation number and the details they need before arriving.

Checkout with guests, promo code, gift card, and itemised visit summary
Fig 09 — Checkout: guests, notes, promo code, gift card, and an itemised summary per guest
Booking confirmation with itinerary by guest
Fig 10 — Confirmation: itinerary by guest, reservation number, and pre-visit information

From a brittle flow to a foundation built to scale

The redesign has shipped. It's too early for reliable post-launch numbers, so here is what changed against each problem I set out to solve.

I'll judge it by measures tied to those problems: booking completion through the flow, completion on mobile specifically, and how quickly a new property can be themed and launched.

Designing the rules, not just the screens

Designing for hundreds of properties meant I couldn't pick the right answer for one business. I had to define a system of defaults, constraints, and flexible parts that gives every property a good answer. That shifted a lot of my time from pixels to decisions about what should be configurable and what should stay fixed.

Talking to clients early made the biggest difference. Hearing how differently they run their properties, and what look they wanted, is what convinced me that adaptability had to be the foundation of the design rather than an add-on at the end.

Back to

All case studies

View all case studies