Case Study: Transforming Personalisation

Challenge
When I joined the personalisation team at On the Beach, the challenge was clear but overwhelming: “fix personalisation.” The team had been tasked with tackling everything at once, creating a huge, unmanageable remit that lacked focus and measurable outcomes.

Approach
In my first two weeks, I focused on understanding the team’s dynamics and the blockers slowing progress. From there, I worked with the team to make some key changes:

• Focused the goal: Broke the monolithic “personalisation” task into a specific, achievable target - building a machine learning hotel prioritisation platform.

• Refined team rituals: Shifted stand-ups from 9 to 10 and moved to every other day, giving the team more time for deep work.

• Brought in OKRs and metrics reviews: Ensured every two-week cycle ended with clear evaluation, iterating when work delivered impact and pivoting quickly when it didn’t.

Solution
We migrated from a static, rules-based personalisation system to a machine learning algorithm powered by over 22 million data points. This algorithm prioritised hotels based on user needs and preferences, while balancing business profitability and contractual agreements with hoteliers.

Impact

• Reduced filter usage on the website by 5.3% - showing customers were finding what they wanted faster.

• More than doubled abandoned basket email click-through rates from 2.31% to 5.7% by feeding them with personalised recommendations.

• Increased productivity and morale by giving the team a more focused, manageable way of working.

Lesson Learned
Fresh eyes make all the difference. Coming into a team with a new perspective allowed me to spot opportunities others had become too close to see. By reframing the problem and adjusting how the team worked, we delivered a better product for customers and stronger results for the business.