This project was created and worked on during the Startup Weekend Brussels 2018 hackathon event. I pitched this business idea and formed a multi-disciplinary team who worked together with me to develop this business idea from concept to prototype.
Travelling is fun but the preparation for it can be hectic. Most people browse through multiple websites to obtain the most accurate information and advice on what to visit, as well as practical arrangements when they travel. Also, most advice obtained on the web is not tailored to each person’s needs and preferences, which adds more complexity into trip planning. Last but not least, when travelling with partner or friends, the preferences of everyone must be taken into account to have a good travelling experience.
What if there would be a smart way to plan your travel, taking into account your preferences and the ones of your friends, saving you time and effort?
Methodology and Solution
Before coming up with a concrete solution, we spent time understanding user needs. I conducted 5 in depth user interviews and collected quantitative data via a survey on travel preferences and time spent. Some major patterns started to occur:
- Most people spend at least 4 hours searching the web for information to prepare their trip
- Most common travel frequency is 1-3 city trips per year
- Only 4% of people travel alone
- There is a need for visiting both touristic places but also for special experiences
- People lacked end-to-end view on their travel planning and cost and wanted to consolidate all information at one place
- A lot of travellers experienced stress from the feeling that they need to maximize their experience but also from the conflicts that can arise with friends or their partner
After having a clear view on user needs, we proceeded to prototyping. The key ideas we wanted to prototype were:
- the possibility to add your travel preferences
- the display of a personalized travel plan
- the possibility to adapt/customize further this plan
Here are some of the draft mockups we created during the hackathon:
Users would start using the interface in a conversational way (inspiration from Netflix) and indicate their preferences, as well as who they travel with and when.
Travelfriends would then generate a recommended end-to-end travel plan, from the moment the person leaves till the return back home. It would include flights, transportation, hotels, experiences, etc, with the possibility to adjust elements the traveller wants to change (e.g. change to a different hotel).
With the use of AI, Travelfriends would learn from people’s behavior and adjust their preferences on the go, making more intelligent suggestions the more the user uses the platform. Thanks to the data aggregation and advanced calculations, travelfriends would predict accurately arrival times, as well as adjust the planning e.g. to avoid busy hours. And when travelling with more people, the suggestions would be customized to consider all people’s preferences (provided everyone is using the app).
Challenges and Conclusions
The key challenge in this project was the access to the data from different platforms (e.g. flights, hotel websites, etc). As different platforms are based on different technology stacks, but have also not made their data public via APIs, it would be a big challenge to aggregate the data. Some possible ideas we explored were:
- Work and exchange data with a limited list of partners: starting with a limited list of hotels websites, flights, etc and exchange data daily based on XML feeds. This is similar to what other aggregators and comparison websites use. The advantage would be the access to all data, but the disadvantage is that the key idea behind travelfiends, to give the best plan based on all possible solutions would be compromised.
- Base our data on Google APIs: as Google has access to a lot of data this was the idea we were considering the most for an MVP version, in order to test the product market fit with the least friction and effort possible. The disadvantage in this solution would be the recurring cost every time we call the Google API to fetch data, which could endager the profitability of the product.
We also considered pivoting the concept to target a specific user group, and provide high-end experiences only. This would make it easier to work with a list of limited suppliers for who we could also guarantee the quality thereof.
Considering the user research results, I strongly believe there is still a high need for personalized travel experiences. It will be challenging to provide a single best experience which fits each person’s needs, but the key is to allow platforms to get smarter and learn from users’ behaviors.