Sankalp Joshi | Rocketrip Recommendations

An AI powered flight and hotel recommendations engine

About Rocketrip

Rocketrip is an enterprise traveler rewards program. Rocketrip offers travelers 50% of the total savings on a trip by booking under the Price to Beat.

For example, Jane is a corporate traveler. She books a flight for $500 while the Price to Beat on that flight is $800, so Jane saved $300 by booking below the $800 Price to Beat, thereby earning $150.

Problem

In an effort to prove value of its product, Rocketrip wanted to drive adoption and increase captured spend from 65% to 75% and push for higher gross savings for the top clients thereby increasing their ROIs.

Role & responsiblities

Led the discovery & design process- user research, customer interviews, ideation, information architecture, wireframing, creating mockups and prototyping.

Launched the 'User feedback panel' - a traveler outreach program to facilitate generative user research. For this initiative, I teamed up with the customer success and support teams for recruiting travelers into this program.

Collaborated with different stakeholders by sharing learnings & insights from discovery, brainstorming different strategies and solutions.

Discovery

Setting up user feedback panel

I launched a traveler outreach program, 'User feedback panel', to interact with travelers. During this project, I teamed up with the customer success and technical support teams for recruiting travelers into this panel. I used this panel to interview travelers to understand their booking behaviors, motivations, booking process and pain-points.

User research & traveler interviews

Interviewed corporate travelers (end user persona) to understand their booking behaviors, motivations, factors affecting their booking decisions, workflows and pain points.
Studied some of the best practices for effect of social norms on user behavior.
Reviewed the in-page analytics and mapped user flow patterns.
Reviewed the in-page analytics and mapped user flow patterns.

Target areas identified from interviews

Safety: All female participants in the study mentioned safety as priority while almost none of their male counterparts mentioned it. Women used information about hotel neighbourhoods, hotel reviews and colleague feedback.
Loyalty programs: Travelers use their own loyalty programs for business travel to get upgrades and earn points.
External reviews: Travelers searched outside of Concur for checking the ratings and reviews of hotels. Some travelers also checked pictures of hotel interiors, rooms and amenities on other travel sites.
Social norms: Travelers showed higher probability of selecting hotels recommended by their colleagues.
Unknown earning potential: Travelers don’t know how much potential they have to earn $ on a trip.

All women travelers described safety as their top priority while no male interviewee considered it as a concern.

Booking experience on Concur

Concur's booking flow is very complex. It requires the traveler to input at least 6 fields upfront to start searching flights. Each search results into 10-15 pages of options.
For example, a flight search between New York and San Francisco will show an average 200 flight results and an almost the same no. of hotels with each hotel offering an average of 5 different room options.
Therefore, travelers on Concur undergo,
- Cognitive overload
  
  Travelers need to know and enter a lot of information upfront before searching.
- Decision fatigue
  
  Endless options and combinations leaving traveler confused.

Solution

Based on the learnings of user interviews & research, we created a recommendation engine. This would create hotel recommendations that travelers are more likely to select. Thus earning more points for the travelers and save cost for the company. A win-win for both!