Project: conversion optimization
Tenzing Travel is a Dutch travel agency offering tailor-made trips to remote destinations. Customers of Tenzing Travel can request offers online by filling in a form on the website. Once an offer for a certain trip is requested, the employees of Tenzing Travel starts acquiring information on ticket prices, hotel availability etc. Once the employee put together the perfect trip, the customer is presented an offer. The customer can choose to book it or not.
One of the challenges of Tenzing Travel is that even though they put quite some time and energy working on these leads, however, only 25% of the offers are being booked. Also, during the months January and February its peak season for Tenzing Travel– everyone starts booking for the summer holidays – and Tenzing Travel receives so many offer requests that not all of them can be responded to.
Because many leads do not get booked in the end, it is crucial to prioritize leads that will be booked. It is difficult to determine which lead will be booked and which one won't. Mostly, this decision is based on gut feeling. During peak season, it is critical to select the right leads.
Gain insights on the conversion of leads and to build a Machine Learning model which predicts the conversion probability of incoming leads.
Using the insights on the conversion allows Tenzing Travel to identify areas in which they can improve. Furthermore, these insights help to test certain hypothesis within the company, for example, is it really better to reply to customers as soon as possible?.
Using data from clients, online offer requests and the historical bookings a Machine Learning model is trained and implemented in the daily operations of Tenzing Travel. This model identifies 'high-potential' offer requests based on the historical booking behaviour.
After implementing our solution, the average monthly conversion increased by about 100% (see Figure 1). The increase in conversion resulted in an estimated increase in revenue of 500k compared to previous year. The model we implemented has played its part in this increase.
Tenzing Travel uses our custom build solution to gain insights into their incoming leads. Delph developed an overview which displays all incoming leads and their information. Every lead is assigned a score that indicates the probability of booking.
The conversion dashboard provides insights into the overall performance of the company and statistics per department, destination and employee are available.
Figure 1. Number of incoming leads and their respective conversion. The green line indicates the implementation of our model.