For companies interested in measuring customer satisfaction, the Net-Promoter-Score (NPS) is a widely adopted method and standard. So what is NPS? NPS could be calculated based on the average response of a set of customers to a specific question: “How likely would you recommend our company to your family or friends?” While several variants of the rating scale exits, the most often adopted version uses 0 for least likely to recommend and 10 for most likely to recommend. Naturally, companies are most interested in customers who express high NPS values (e.g. 9 or 10) that are referred to as promoters and customers who express low NPS values (e.g. less than 5) that are referred to as detractors. If we could understand the reasons that turns a customer into a detractor, we could improve these aspects of the product and increase the quality of the service. Similarly, understanding what makes our customers become promoters allows us to further enhance these aspects of the business and create more loyal and satisfied customers.
HOW DO WE COLLECT NPS FEEDBACK?
On the hotel search results page on the Hotwire site, the customer recommendation score plays one of the most important roles in terms of informing the potential customer about the quality of the displayed product.
This recommendation score is collected by asking our customer who booked their travel to the specific hotel via Hotwire to fill out a simple survey. One of the optional question on the survey form is the above mentioned NPS-focused question. Despite the fact that not every traveler submits their reviews and out of those not all respond to the NPS question, Hotwire was able to collect a set of several millions of NPS responses over the past couple of years. When we joined this NPS feedback data provided by our customers with additional information describing the specific hotel they purchased as well as data describing the search they performed, we have created a perfect data set to try to answer the above questions of what turns one customer into a promoter and another into a detractor. Is it the amount of discount that customers get by shopping at Hotwire? Or is the quality of the product and the travel experience they purchased?
WHAT CAN WE LEARN FROM THE DATA?
To answer this question, we have interpolated the collected NPS data onto a 2D grid. The x-axis and y-axis of the grid expressed the values of displayed savings percentage and the recommendation percentage of the hotel that the customer booked before they submitted their NPS score feedback. The individual cells on the grid were populated with the average values of the NPS survey response in the collected data set. The below figure shows the result of this visualization:
In the figure, the dark blue color corresponds to NPS values of 2 (i.e. strong detractors) and dark orange color corresponds to NPS values 9 (i.e. promoters). As can be seen in the figure, the main driver for the NPS score appears to be the recommendation percentage, where values below 30% recommend were most frequently resulting in customers turning into detractors while values above 90% were the key to turning a customer into a promoter. The figure also clearly depicts that in addition to the recommendation score, the amount of money customer saves on the Hotwire site plays additional role in driving the number of promoters higher. As a result, quite intuitively, we can observe most of our promoters to be customers who purchased above 90% recommended hotels with above 40% savings. The Hotwire Data Science team incorporated these insights into the current hotel search results sorting algorithm that we will describe in another blog post.