Navi

Personal project

Year: 2024

Problem definition

Although there are countless apps and websites for travel planning, many travelers feel overwhelmed when organizing their next trip. The abundance of suggestions from different sources makes it difficult to identify what truly matches their needs and preferences. Planning a trip goes beyond simply choosing destinations and activities, it also involves considering time constraints, advance bookings, suitable accommodation, and dining options that fit both budget and taste.

Solution

Navi is a travel app that uses AI to generate personalized routes, helping users plan their next destination with ease. It provides reviews, booking details, and all the essential information needed to create a smooth and tailored travel experience.

Competitor analysis

A competitor analysis was conducted to evaluate existing AI-powered travel planning tools. Even the most advanced solutions (e.g., Tripadvisor’s AI planner) face challenges in adapting suggested itineraries to real-world conditions such as timing, mobility, and other contextual factors. Moreover, a study by Cornell University analyzing 500,000 real itineraries found that fewer than 10% of those generated by state-of-the-art language models reached ‘human-level’ quality in terms of feasibility, rationality, and customization.

First prototype and user test

The first prototypes consisted of sketches and low-fidelity Figma wireframes aimed at defining user flows, information architecture, and page navigation. I conducted 23 user tests with participants who matched the target audience and had varying approaches to both travel planning and the use of AI:

  • 42% preferred to meticulously plan every detail in advance, while 58% preferred a more spontaneous “go with the flow” style.

  • 74% regularly used AI tools like ChatGPT in their daily life and work, reporting a relatively high level of trust (rated above 6 on a 1–10 scale). The remaining 26% rarely used AI tools and reported lower levels of trust.

Participants completed a set of tasks using the think-aloud protocol.
The tests revealed several key insights:

  1. Users valued having a personalized route but also wanted inspiration from other travelers’ itineraries. On the discovery page, they expected not only destinations and activities but also sample routes created by others.

  2. During the initial survey, users wanted the option to select more than one destination.

  3. Not all trips were leisure-focused. Some users had limited time due to work or fixed appointments. It was necessary asking for the available free time during the initial survey.

Final prototype

The insights from prior research guided the design of an intuitive and seamless user experience. The interface was designed to be simple yet sophisticated. Neon green is used as the accent color, while a neutral palette of white, black, and grey maintains a clean, balanced look. This choice allows photos and maps to stand out as the central elements of the experience.

Here are explained some of the main sections of the final prototype.

1. Registration and Access

2. The steps to generate a new route

A central button in the navigation bar invites users to create a new route. Tapping it starts a quick quiz that leads to personalized travel suggestions.

The quiz covers key details:

  • Destination
  • Travel dates
  • Daily availability
  • Budget
  • Travel companions
  • Interests

The final summary lets users review and edit their answers before confirming. The app then generates personalized routes based on their preferences.

3. The generated routes

After completing the survey, users receive personalized travel options tailored to their preferences. They can further refine their itinerary by adjusting the timeline or selecting different destinations.
Each route includes a brief overview, suggested activities, recommended hotels and restaurants. Users can choose to view the itinerary as a timeline or on an interactive map that highlights destinations and ideal visiting hours.

Every destination provides key information at a glance: a short description, opening hours, website, and ticket booking options. Additionally, users can read travel tips and insights shared by other travelers who have previously visited the location.

4. Community & Local guide

The community aspect plays a key role in the travel experience. It’s something AI can enhance, but never replace.
Each user has a personal profile showcasing the countries they’ve visited and their custom itineraries. Users can also become local guides, sharing hidden gems, cultural insights, and local traditions with travelers.

This feature helps foster meaningful connections between tourists and locals while promoting lesser-known destinations and creating new opportunities for local communities.

Final thoughts and metrics to be measured

Navi is a concept app I designed to demonstrate my approach to mobile product design and to highlight my interest in working on similar projects for real clients.

If launched, I would track key metrics to measure success, including:

  1. Acquisition (Number of downloads, Cost per acquisition, Traffic sources)
  2. Activation and onboarding (Activation Rate)
  3. Engagement (Daily Active Users (DAU) / Monthly Active Users (MAU), Stickiness Ratio (DAU ÷ MAU), Session Length, Session Frequency)
  4. Retention & Churn (Retention Rate, Churn Rate, Uninstall Rate)
  5. Performance & UX ( Crash Rate, App Load Time, User Ratings & Reviews, Support Requests)
Contact

Are we a good fit?

Great design comes from working closely together. That’s why I like to make sure we’re a good fit for each other. Let’s have a quick call to see how we could work together.