Conceptualising Covid-19 recovery strategy

Problem

  • As a part of the company-wide Covid-19 response, my product team worked to make it easier for guests to find more spacious places to stay that offered high levels of privacy.

  • The primary objective was to help users discover, filter, compare and book places to stay where they would feel comfortable and safe. This required defining levels of privacy that were scalable to all property types. 

  • I was tasked by my senior product manager to lead a cross-team alignment session to kick-off the project with a content-first approach. My goal was to concretise the abstract concept of privacy levels and to ultimately set the stage for defining and scaling this product line.

My deep dive in preparation for the project kickoff (read top to bottom, left to right).

My role

  • Did a deep dive into the concepts of space and privacy, exploring what makes something more or less private, tangible indicators of ‘levels of privacy’ and how we could map these levels to all combinations of unit and property types on Booking.com.

  • Uncovered and discussed a clash between perceived and actual privacy: it is possible to book a seemingly stand-alone villa (unit) at a not-so-private hotel (property). 

  • Created a positive space to discuss the ‘why’ behind our work with designers, developers and product managers which built the necessary alignment for what turned into a multi-year line of experimentation and product development.

  • Conducted rounds of usability and A/B testing that unequivocally showed the value of emphasising unit information over property information, something that was previously not appreciated within the company.

These are examples of listings where the content hierarchy now includes information about privacy.

Outcomes

  • Using the rules and content standards defined in our kick-off, the team successfully added privacy-level information to all listings on Booking.com, displaying it across app, mobile and web platforms. This led to a huge win on already heavily-optimised placements, delivering a 1.07%, 0.63% and non-inferior increase in bookings, respectively. 

  • My product team ran many other experiment lines to emphasise privacy at a unit level in order to optimise impact. A checkbox we added to allow users to filter for entire units drove 3,000 additional bookings per day.

  • I delivered two significant content ‘wins’ on the Booking.com search results page: reframing a sorter (which increased bookings by 0.21%) and repositioning a promotional banner (which increased bookings by 0.5%). 

  • Building upon the success of emphasising units over properties, I turned my attention to improving the names of our units: I changed ‘chalets’ to ‘cabins’ in the US in order to begin speaking our users’ languages more accurately (which led to a 12% increase in engagement with the unit type in the US).

  • Our product team defined a new way for the company to measure impact: a unit-based booking metric for our experiment tool.

The checkbox which allows users to filter for 'entire places' at the unit level.

The sorter, which filters for homes and apartments, now emphasises space and privacy.

The promotional banners which target specific audiences, now appealing to the post-covid traveller's needs.

 
 

A few example placements where I changed ‘chalet’ to ‘cabin’, including filter name, unit name and a USP.