Nobody arrives at a travel site knowing the destination, they arrive knowing the feeling they want. So discovery is organised by mood, not geography. Five moods, each with a plain-language line and a live count: Urban Escape, where skylines never sleep; Nature & Tranquility, where silence speaks; Cultural & Historic Wonders; Nordic & Dramatic Landscapes; Timeless & Immersive Asia. Selecting one plots its destinations on the atlas and zooms to that region. Thirteen journeys become a choice of five, and the map does the narrowing.
04 · DESIGN HOLIDAYS · 2026 · ROLE: SOLO DESIGNER + BUILDER
A brochure trapped in PDFs, rebuilt as a way to choose a journey.
- Situation
- Design Holidays had an old website in shambles and a product that really lived in fragmented PDF files, one per travel plan. They wanted a quick, credible entry back into the market.
- Clarity
- I designed the whole thing solo and built it with AI assistance, turning scattered plans into one system: mood-led discovery on an atlas, and a single repeatable journey template behind it.
- Shipped
- A live eighteen-page site, five core pages and thirteen journeys, with mood-based discovery and WhatsApp-first inquiry, delivered in under a month.
01 · The mess
The client's old site was doing them no favours, and their real product lived in PDF files emailed out one at a time.
They needed to be back in the market fast. But the problem wasn't decoration. Anyone shopping for a holiday has already seen ten identical sites, the same grid of destinations, the same stock photos, the same "book now". Nothing in a list of thirteen places tells you which one is yours.
And the business actually ran on WhatsApp, not a checkout. Building a conventional shop would have modelled the wrong thing entirely.
- 01Old site in shambles
- 02Plans trapped in fragmented PDFs
- 03Thirteen destinations, no way to choose
- 04Every competitor looks identical
- 05Business runs on WhatsApp, not checkout
- 06Quick market entry needed
02 · The work
Research
The whole brief was a stack of itinerary PDFs. There was no site to improve and no design system to inherit, so the research was competitors: what patterns travel buyers already knew, and which of them were worth borrowing rather than inventing.
- Studied competitor travel sites for the conventions customers already recognise
- Chose the familiar pattern over the novel one wherever both would work, speed to market was the actual requirement, not originality
- The client already had the marketing channels; the site had to be ready to receive them
Collaboration
One document in, one site out, and everything in between was mine, with the founder as the only reviewer.
What I was given
Itinerary documents
Day-by-day text, hotels, inclusions, exclusions. A logo from the client, and nothing else, no site, no reference.
What shipped
A live booking site, on their own domain
Thirteen journeys through one template, on a domain they already owned and already marketed.
Everything between the two, mine
- Research
- IA
- UX
- UI
- Copy structure
- Build
- Ship
The logo came from the client. Everything built on top of it is mine, and iterations went straight to the founder and came straight back, which is the shortest review loop on this site and the reason it reached the market quickly.
03 · The big calls
The client converts over WhatsApp, so the design makes WhatsApp the primary action instead of burying it under a form. Every journey page carries a sticky booking card: WhatsApp first, email second, and a full inquiry form for people who want to write more. Pricing sits in the open in rupees on the fact strip rather than behind a 'request a quote' wall, a boutique operator competing on service should not make the first interaction an obstacle.
I designed in Figma and Photoshop and built the site with AI assistance, which is why a solo project of this size shipped in under a month. That speed only worked because the design was systematised first: one journey template, hero, sticky section nav, fact strip, highlights, day-by-day accordion, inclusions against exclusions, inquiry, filled thirteen times from the client's PDFs. The template was the real deliverable; the thirteen pages were data.
04 · The build




1 · Sold on feeling, not geography, the premise, above the place names.
2 · Count visible before the click: one destination is never a surprise dead end.
3 · The map answers by subtraction, only the chosen mood is plotted.

- 1 ·Sold on feeling, not geography, the premise, above the place names.
- 2 ·Count visible before the click: one destination is never a surprise dead end.
- 3 ·The map answers by subtraction, only the chosen mood is plotted.
From a feeling to a conversation
A mood
five ways in, a feeling, not a destination
The atlas
plots only that mood, zooms to the region, states the count
One journey
duration, location and price in the open
the primary action, where the business already talks
Email or inquiry form
for people who want to write more
No checkout, the conversation is the conversion

Problem → part
- Fragmented PDF plansOne repeatable journey template
- Thirteen destinations in a listFive moods on an atlas
- Generic destination gridDiscovery by feeling, not geography
- Hidden 'request a quote'Open pricing, WhatsApp-first inquiry
- Old site in shamblesLive eighteen-page site in under a month
05 · The facts
★ Ryan Soni
Case 04 · DESIGN HOLIDAYS · 2026
<1 mo
Design → live
13
Journeys shipped
Solo
Design + build
- Client / domain
- Design Holidays, a boutique travel studio in Chennai, India, selling curated international journeys
- Live site
- designholidays.in, public; the client cleared this case study
- Timeline
- May 2026 → shipped in under a month
- Role
- Solo designer and builder, end to end
- Team
- Ryan solo, working directly with the client
- Method
- Designed in Figma and Photoshop; built with AI assistance (Antigravity). The design decisions, system, and craft are mine; the AI accelerated the code.
- Scope
- Five core pages and thirteen journey pages, brand application, design system, page design, responsive behaviour, and the front-end build
- Starting material
- Fragmented PDF files, one per travel plan, plus an old site in poor shape
- Design system
- Editorial serif display against a sans body, cream ground, ink navy, brass accent, chosen to read as a boutique studio, not a booking aggregator
- Discovery model
- Five moods → atlas pins with live counts → journey detail → WhatsApp or form inquiry
- Accessibility
- The mood selector is a real tab set (role="tab" / aria-selected), with a desktop rail and a scrollable mobile pill bar
· Outcome ·
The client reports that marketing drove a high volume of clicks and that business is running ahead of what they anticipated. No figures were shared with me, so none are claimed here.
None, the client approved public use of this work
06 · The takeaway
This is what I do when there is nobody to hand off to. One product call, choose by feeling, not by map. One business call, talk to people where they already talk. Live in under a month, because I built the system before I filled the pages.


