TravelMesh, Web Product Design
Tourism Services Booking System
TravelMesh is a web product of a tech startup company from Trento, Italy. It is a web based platform that connects existing online tourism booking systems through a meta-search technology and outputs the travel offers in a UI web widget and an API. TravelMesh enables travellers to search, create and book custom travel packages formed of local tourism services, otherwise available on multiple disconnected websites and systems.
In two years, starting fall 2013, I lead entire product design process starting from product idea, through research, prototyping, investor pitching up to building the Minimum Viable Product (MVP), branding design and user testing. Since 2014 TravelMesh was installed on various destination websites, ranging from national scale, such as Expo 2015, with millions of monthly visits, down to local scale destination websites in Trentino province.
Year: 2014-2015
Startup: TravelMesh
Role: Head of Design, part of the in-house agile development team
Responsibilities: Interaction, Branding and Visual Design, User Testing and Research
Design Tools: Illustrator, Photoshop, InvisionApp
PrototypesConsumer widgetSupplier Onboarding
PRODUCT BRIEFING
Client commissioned me to research and design a digital product solution that they would pitch to some potential investors. They wanted to address a problem of the organizations involved in travel destination marketing (DMO) who were lacking appropriate digital tools for competing with tech savvy giants such as Booking.com. With their outdated, complex and costly ecommerce management softwares, based on expensive licensing models, DMOs were not able to manage their online ecommerce catalog and their relations with suppliers and customers in a fast paced fashion as dictated by the modern internet market.
Client’s initial solution idea was to modernize and implement current standards for this sector by creating a new better destination management and booking software (DMS):
 - Complete booking solution for multiple accommodation providers
 - Proprietary payment system
 - Back-office area for multi-user availability management
 - Subscription based SaaS multi tenant model
 - Modern, responsive, fast, usable and customizable user interface
I was asked  to deliver research documents and a graphic prototype to back the idea, to illustrate a potential solution and to highlight a business and innovation opportunities.

Research documentation: a 112 pages document containing insights from desktop research and user interviews
Research
I started with desktop research focusing on the current state of the online travel industry, followed with interviews with DMOs, testing of several pieces of software used by DMOs and destination operators, alongside workshops with hundreds of tourism service suppliers and benchmarking of hundreds of existing online booking solutions.


MY RESEARCH CONCLUSION
Initial idea to create a cutting edge software had many competitors, would be too costly and difficult to market and probably would be not addressing real problems and opportunities.
 - Feature opportunity - travellers were booking all type of tourism services online, thus there was room for testing a possible mashup of services, the so called dynamic package
 - Strong competition - building a better booking software for all type of tourism services could provide some advantage over competition in terms of features (most solutions are focused only on room bookings), but would end up being less affordable
 - High marketing and support costs - suppliers were not showing any willingness to invest their time in updating an additional inventory system
 - No market for a costly solution - the DMOs wanted to compete by focusing on their strengths - branding and destination promotion - and spend less on the online ecommerce

MY NEW SOLUTION SUGGESTION
My advice was to drop the idea of building just another online booking engine in favor of creating a meta-search engine - connecting the already existing booking systems (supplier’s website, online travel agencies, management software, channel managers).
This way, Tourists would still book through the DMO’s website, but directly to the service provider. They would be able to add a room, a tour and a museum ticket to a single shopping cart and pay only once. Meta-engine would take care of the split payment and booking to different providers. Suppliers, on the other hand, will not have to bother to update an additional inventory, they will connect and use their existing favorite system. By putting the traveller directly in relation with the supplier, DMOs would cut on costs of bookings management.
This idea was following current trends, lacking competition and there was space for innovation.
DESIGN CHALLENGES

Business, marketing and branding challenges were on hold, the strategy was to get the Minimum Viable Product (MVP) out in the wild in order to test and track data from a real world users. Timeline was tight, the beginning of the tourism summer season was less than six months ahead. My responsability was to design a prototype and make sure that the engineering team’s work and decisions were following the big picture.
The most challenging part for me was designing for several groups of end-users: Store owner (DMO), Supplier (Hotelier), Customer (Traveller). This is a typical online middleman ecommerce story. I had to design touch points for every user type while considering their crisscrossed experiences, risks and challenges. There was also technical providers user group: Data provider (Booking Engine), Payment systems (Online bank), Online store builders (Web design agency).
Before designing the user interface, we needed to connect existing booking systems APIs. Normalizing data and the UI was an iterative process. Often I would get back to design when a new data set was connected. I was focused on creating an universal UI and API standard for each kind of tourism service, instead of adapting to each service provider.
Wireframes help fast iteration - simple black and white illustrations, created first on paper and then digital in Adobe Illustrator, were uploaded to InvisionApp in order to secure collaboration of everybody involved in the product definition.
Using iframe widgets to integrate the UI with any CMS was limiting. I worked closely with front-end developers to ensure that the best possible solutions are applied. Modal windows, payment security, deep linking and responsive frameworks were challenging with iframes.
There was need for flexible but stable visual and UI design as every widget would get custom CSS skinning in order to fit the hosting store’s website design. My concern was to standardize UI elements and make room for a right amount of customization, but in the same time guarantee clarity, usability and conversion rates.
Product design process: from sketches to an online collaborative wireframe to a final visual prototype.
Sketching and visual design thinking is fundamental part of my designing process. Most decisions take place at this stage. Usually pen on paper, iPad in travel.
DESIGN CHALLENGES AND PROCESS
Before designing the user interface, we needed to get in touch with the real data through a live booking system. Lodging booking engines were lacking standards, thus we had to first normalize the data. Fortunately, during my research I came across an international project working on open standards for the travel industry. We adopted and even contributed to the standards.
Designing a service booking interface was an iterative process. It was applied to the design of each type of tourism services. Often, I would get back to design when new data set was connected. In this way we were  able to focus on creating an universal UI and API standard for every kind of tourism service, instead of adapting to each service provider. There are services which are availability based (hotel rooms, tours), others, having an inventory, are more similar to physical products (tickets, rentals), while some do not have capacity logics (tourist cards, ski pass). 
Using iframe widgets as the simplest UI integration method was limiting possibilities of the UI design. I worked closely with front-end developers to ensure that the best possible solutions are applied. 
There was no need for visual design as every widget would get CSS skinning to fit the hosting store’s design. Default visual design was based on Bootstrap framework’s elements. What I focused on was standardizing UI elements that would get right amount of customization in order to resemble any visual identity. 
In the tourism industry one of the top priority features is multilingualism. My experience in designing multilanguage products and my good knowledge of three languages helps me when designing and working with translators. As text is one of essential elements of UI, my concern is to understand, research and test translations.
Widget UI visual customization - SASS, Admin UI, Personalized front-end Booking widget
Live product
Since the launch in 2014, TravelMesh gathered clients ranging from big public funded websites, such as dedicated website for the Expo2015 in Milano (1000+ suppliers) to small local private associations, such as Andalo Vacanze with 40 hotels. Working closely in touch with clients and their customers secured an user centered approach in product's evolution.
Campiglio Dolomiti Official Madonna di Campiglio Ski Area tourism website TravelMesh widget
EXPO 2015 Milano, Official Tourism Website - Checkout process
Garda Lake Trentino, Official Tourism Website - Lodging Results Page
Garda Lake Trentino, Official Tourism Website - Product Results Page
City of Parma, Official Tourism Website - Upselling modal screen
CONVERSION RATE IMPROVEMENT
Millions of tourists searched and booked through TravelMesh UI widgets and APIs. This was a great opportunity for data backed design decisions. We used different systems for data tracking and analysis (custom server data log for APIs, Google Analytics for visits and Heap Analytics for UI and user flow tracking). Data was analysed manually, hence we could work on improving the conversion rate. The average conversion rate improved by 33% in 6 months, thanks to constant analysis and UI improvements.
The average conversion rate from the product list to completed booking improved by 33% in 6 months, thanks to constant analysis and UI improvements.
NEW FEATURE CASE
Some features were developed upon explicit requests from store owners, from their suppliers or from usability tests. Others were improvements developed based on the results from data analysis. For example, while analysing a conversion funnel, I spotted that the hotel detail page had a high number of exits. Tourists would search for availability, identify a potentially interesting hotel and then open the detail page for additional information (photos, services, location). The next step would often result in an abandonment of the booking process. Assessment showed that not being able to proceed straight to the booking from the hotel detail page was a blocker. I designed and worked closely with developers to deliver rapidly the new feature: query and book an individual hotel directly from the detail page. It improved the conversion rate by 50%.
Enabling booking of an individual Hotel, without querying the entire inventory, required major changes in the API meta-search structure.
ADMIN AREA DESIGN
When you imagine a subscription based commercial digital product, you figure some sort of admin area that requires login and comes with all kinds of settings pages. This became my priority after the business plan and the working strategy was defined.
Administration console for DMOs, the most valuable client group: reports, insights of the widget performance, booking details, lists of suppliers;
Some of the major challenges was the multitude of user levels resulting in a cascade of dependent permission levels, which needed to be defined and controlled.
DMO Admin - Admin area login
Supplier onboarding
Registration and terms acceptance by tourism service suppliers was strategical. As the average user resulted as “not much tech savvy”, we needed the simplest possible onboarding process.
BRANDING AND MARKETING
From the very start of this project; I was involved in the strategy meetings, marketing decisions and business planning. I have designed key elements of the product brand - the name and the visual identity. Brand is important part of the product, being usually at the very start point of the user’s product experience. 
I have designed and produced all marketing material, from digital assets (website, blog, newsletter, email signature, presentations, templates) to physical (business card, brochures, banners).
CUSTOMER SUPPORT
As customer support is an important part of the product’s life and marketing, I have spent significant time replying to tickets and responding to telephone calls. It gave me the opportunity to develope and design material for a better customer service (guidelines, instruction brochures, technical documentation, newsletters).

Visual brand mood and assets - modern look for a tech startup
Designs for Newsletter, PDF brochure, Blog
TravelMesh website, Dynamic Packaging feature page (automatic translation).
Design and copywriting for the technical documentation and installation guidelines.
PRODUCT DESIGN CONCLUSION
The idea of solving the problem through a facilitator metasearch platform, which could connect to any kind of existing booking system, proved efficient. DMOs and Service Suppliers are enthusiastic with having this possibility, as no other platform at the moment offers such a service. It is a real opportunity to bring online all local services of a tourism destination, without the need for the suppliers to acquire or adopt any additional software.
“With TravelMesh we got rid of management costs: we no longer need to update the availability of the services … we are free to focus our energies and our financial resources on marketing and online promotion.”
- Andalo Holidays
“I'm really glad that EXPO 2015 contributes to the TravelMesh experiment which undoubtedly has great elements of originality and innovation.“
- EXPO Explora DMO
“TravelMesh merged very well with the new design of our website. The previous system was outdated and too expensive to adapt.”
- Campiglio Dolomiti